I create content about deep learning | The DevRel and Deep Learning Dude

I don’t wanna hop on call to learn about or see a demo of what you’re building. Too busy hacking on VLMs, curating multimodal datasets, and contributing to the open source community. 

Harpreet Sahota 🥑

It was great to be back in Boston last week and host another awesome meetup!

Thank you for everyone who showed up, our awesome lineup of speakers for taking time out of their schedules for delivering such informative talks: Anku Rani Hanhui Wang Sage Simhon Sharut G. 

And for Voxel51 for sponsoring!

If you missed the event, checkout the comments for where you can find the recordings of the talks

🤖 Hacker-in-Residence @ Voxel51| 👨🏽‍💻 AI/ML Engineer | 👷🏽‍♀️ Technical Developer Advocate | Learn. Do. Write. Teach. Repeat.

No alternative text description for this image

i built a ComfyUI-inspired visual, node-based workflow editor for FiftyOne

now you can build dataset curation, analysis, and model evaluation pipelines by connecting nodes on a canvas, without writing code

give it a try and let me know what you think, completely open source and all you have to is `pip install fiftyone` and follow the instructions on the readme

everyone is hype about gtc, but the 3d vision conference is this week and i'm much hype for that

i made a repo and a dataset to explore the papers

checkout the repo here:
 https://lnkd.in/gRFz6Kb7

here's a dataset that you can use to explore the papers:
 https://lnkd.in/g2RA87fD

it's been a long, long while since i've done a live stream on linkedin, but it's happening in t-mius 8 minutes!

if you're interested in learning about how to work with video datasets using open source tools, open data, and open models...this is the session for you!

Join us on Feb 26 for Exploring Video Datasets with FiftyOne and Vision-Language Models. Register for the Zoom: https://lnkd.in/gCk7gGPi

Video is the hardest modality to work with — more data, more complexity, and annotation workflows that don’t scale. This hands-on session by Harpreet Sahota 🥑 shows how to understand what’s inside massive video datasets without manually watching thousands of clips.

You’ll learn how to:
🎬 Navigate and explore video datasets in the FiftyOne App and understand dataset structure
🔎 Compute embeddings with Qwen3-VL to enable semantic search, zero-shot classification, and clustering
📝 Generate video descriptions and localize events using vision-language models like Qwen3-VL and Molmo2
🧭 Visualize patterns and trends through embedding projections in FiftyOne
📏 Evaluate model outputs against Action100M’s hierarchical annotations

Perfect for ML engineers, data scientists, and computer vision teams working with large-scale video data who want practical workflows for exploration, curation, and debugging.

#computervision #machinelearning #ai #datascience #videoanalytics #visionlanguagemodels #mlops #opensource

Derek Van de Streek, Swati Bannore Jain and 11 other attendees

Feb 26 - Exploring Video Datasets with FiftyOne and Vision-Language Models

excited to be back in stuttart! if you're in the area and want to speak at the meetup on april 21st, send me a dm

Chief Community Officer @ Voxel51 | Startup Advisor

Are you based in the Stuttgart area? Consider speaking at the upcoming April 21 - Stuttgart AI, ML and Computer Vision Meetup: https://voxel51.com/speak

#computervision #ai #artificialintelligence #machinevision
#machinelearning #datascience #opensource #stuttgart

i've learned a lot from Eric Ma over the years, from the first time I saw one of his talks on principled data science workflows to talks he's given for my communities over the years, and his latest work on getting the most from coding agents.

check out his next talk, and support the SciPy community as well!

Together with my teammates, we solve biological problems with network science, deep learning and Bayesian methods.

As a data scientist, how much of your week is spent on work a coding agent could do?

I once planned multiple days of work for a machine learning experiment last year. This was an experiment that I had put off for an entire quarter because, as a team lead, I didn't have the time to block out the 3-4 afternoons that I thought I'd need. 

Instead, I described the problem to Claude Opus 4.5 via OpenCode, and it wrote the code, ran the training, found issues, switched strategies, and finished the whole thing in 20 minutes, complete with a written report drafted *with* figures referenced.

The data scientists who figure this out now will have an edge.

I'm teaching a workshop on how to do this flavour of agentic data science to raise funds for SciPy 2026's Financial Aid Program. You'll learn:

- Prescriptive prompting for experiments
- Project structure for agents
- Logging and verification strategies to enable automated control
- Report writing with AI
- Staying in control during EDA
- The minimalist version first

These are the same techniques I use in my personal projects and my teammates to compress data science timelines from days to minutes. I think you'll have a chance to learn something good while doing something good!

Details:

- Date: March 27 & April 3, 2026
- Time: 2:00-4:00 PM EST
- Venue: Google Meet (calendar invite will be provided)
- 40 spots available
- $1,000 minimum donation, paid directly to NumFOCUS (tax-deductible), student pricing: $200!

Video recordings available if you can't attend live.

Sign up here: https://lnkd.in/eFjc36Tb. I'm looking forward to seeing what you'll build with these techniques! 

P.S. If you can't make it but share it with others who have the means and would benefit from the workshop, I would deeply appreciate it!

Open article: Agentic Data Science Workshop | SciPy 2026 Fundraiser by agent-assisted-data-science.vercel.app

Agentic Data Science Workshop | SciPy 2026 Fundraiser

We just released a new AWS workshop that I’m really excited to finally share 👇
https://lnkd.in/eiNprTd7

This came from something I keep seeing across autonomy, robotics, and computer vision teams lately.

Most teams are not blocked by models anymore.
They are blocked by data and iteration speed.

So we wanted to show what a practical workflow actually looks like:
🚗 generate synthetic driving data
 🧠 auto label using foundation models
 🔧 fine tune a small vision language model
 ☁️ deploy and iterate using AWS SageMaker
 👀 inspect and improve everything inside FiftyOne

The goal was simple.

Build something real that teams could follow end to end.

Huge credit to Ryan Patrick Sweeney, Mark Vinciguerra, and Derick Thompson for driving this forward, along with great collaboration between Amazon Web Services (AWS), NVIDIA Omniverse, and Voxel51 🙌

Also excited that Voxel51 is now live on AWS Marketplace, which makes getting started a lot easier for teams working on computer vision and robotics workflows!

Personally, I think this is how a lot of Physical AI systems will get built going forward.

Faster loops between data, models, and deployment instead of long waiting cycles.

If you’re working on robotics, AV, or multimodal systems, I’d genuinely love to hear what you’re building.

More coming soon 🚀

#PhysicalAI #ComputerVision #Robotics #MultimodalAI #SyntheticData #AWS #MachineLearning #AutonomousSystems

Open article: GitHub - aws-samples/sample-finetuning-small-vlms-synthetic-data by github.com

Contribute to aws-samples/sample-finetuning-small-vlms-synthetic-data development by creating an account on GitHub.

GitHub - aws-samples/sample-finetuning-small-vlms-synthetic-data

I’m so happy to have my boi Wiz present at one of our meets!

AI @ NVIDIA - 🏗️ Build | 🚢 Ship | 🚀 Share

View Chris "The Wiz 🪄" Alexiuk’s profile

skills.md - you've seen it, you've probably used it (I assume if you're following this LinkedIn account, you're someone who uses tools like OpenCode/etc.

I'm excited to be yappin' about skills at an upcoming Voxel51 / Hugging Face virtual event with a line up that boggles the mind: 

Ben Burtenshaw, Daniel Ávila Arias, and AdonAI Vera.

My presentation will focus on something near and dear to my heart: Skills as documentation. 

The days of documentation aren't over - humans, and agents, still need to understand how things work - but they're changing quickly. Not everyone wants to read a thorough getting started guide that may be built for a specific system (and if it's not, it'll be ~2K words long) Sometimes, you just want to clone a repository - and get cooking. 

Skills help bridge that gap - context for LLMs, but tailored for a human experience. 

Can't wait for the event, come on out and ask any skill questions you have!

looking forward to hosting this workshop tomorrow, join me!

Join Harpreet Sahota 🥑 on Feb 26 for a virtual workshop to learn how to use Facebook's Action100M dataset and FiftyOne to build an end-to-end workflow - https://lnkd.in/gFvwbQni

What you'll learn:

* Navigate and explore video data in the FiftyOne App, filter samples, and understand dataset structure

* Compute embeddings with Qwen3-VL to enable semantic search, zero-shot classification, and clustering

* Generate descriptions and localize events using vision-language models like Qwen3-VL and Molmo2

* Visualize patterns in your data through embedding projections and the FiftyOne App

* Evaluate model outputs against Action100M's hierarchical annotations to validate what the models actually capture

By the end of the session, you'll have a reusable toolkit for understanding any video dataset at scale, whether you're curating training data, debugging model performance, or exploring a new domain.

***********

Want to build better computer vision models? FiftyOne is an open source toolkit from Voxel51 (our Meetup sponsor) that helps you curate datasets, evaluate model performance, visualize embeddings, catch annotation errors, and eliminate duplicate images—all in one place.

“pip install fiftyone” is all it takes to get started - https://docs.voxel51.com/

#computervision #ai #artificialintelligence #machinevision #machinelearning #datascience #opensource

I'm passionate about harnessing the transformative power of Artificial Intelligence to solve complex business challenges and create significant, measurable value. I love to lead AI innovation, building and executing strategies that integrate cutting-edge technologies like Machine Learning, Generative AI, and advanced data analytics into the core fabric of an organization.   

With over 10 years shaping data-driven strategies and solutions , my leadership philosophy centers on building, empowering, and mentoring high-performing, cross-functional teams capable of tackling ambitious AI goals. I thrive on setting a clear strategic vision for AI, developing actionable roadmaps aligned with business objectives, and fostering an organizational culture where data literacy and AI-driven innovation flourish.   

My expertise spans the full AI/ML lifecycle, encompassing Machine Learning, Deep Learning, NLP, MLOps practices, cloud AI platforms (focused on Microsoft Azure), and robust Data Strategy formulation. In roles including Global Head of Data at QIMA and Head of Enterprise Analytics at Philip Morris International (a Fortune 500 company) , I've led the design, development, and implementation of scalable AI and data solutions. My focus has always been on leveraging data foundations not just for insights, but as the critical enabler for advanced AI capabilities that drive tangible business outcomes like operational optimization and enhanced performance.   

My background in finance and risk management  provides a strong foundation for ensuring AI initiatives deliver demonstrable ROI and align with core business needs. I excel at bridging the gap between technical possibilities and business strategy, communicating complex AI concepts effectively to diverse stakeholders, from technical teams to C-suite executives, fostering alignment and driving adoption. As an instructor and published author on advanced ML , I am committed to continuous learning and staying at the forefront of AI advancements, with a particular interest in the responsible development and deployment of next-generation AI systems.

Courtlin Holt-Nguyen

You would not hire a chef who has never cooked a meal. So why do so many CS grads expect to land top jobs with nothing but a transcript and a handshake?

I see it every hiring cycle. Smart, ambitious candidates. Zero portfolio. No GitHub. No Kaggle. Not even a personal website. Just a list of classes and a GPA.

Here is the uncomfortable truth: Your degree is the ticket in. Your work is what gets you the seat.

If you want to stand out, you need proof. Not just potential.

Start with one project. It does not have to be fancy. Pick a realworld problem. Build a solution. Document your process. Share your code. Reflect on what you learned.

Repeat. Each project is a story you can tell in interviews. Each repo is a signal to employers: I do not just know the theory. I can build.

If you are still in school, you have time. If you are out, it is not too late. The best time to start was yesterday. The second-best time is now.

I once interviewed a CS grad who had a 4.0 GPA and a resume packed with honors. On paper, he was perfect. But when I asked about projects, he froze. No GitHub. No apps. Not even a class assignment he was proud of.

He told me he thought his transcript would speak for itself. That employers would see his grades and know he was capable.

I wish I could say this was rare. It is not. Every year, I meet brilliant students who have never built anything outside the classroom. They are shocked when the job offers do not come.

I remember my own first portfolio. It was messy. Halfbaked. But it was mine. It showed I could turn ideas into code, not just memorize algorithms.

That portfolio opened doors my GPA never could.

If you are a student, do not wait for someone to assign you a project. Build something. Anything. Show the world what you can do.

Oops, Amazon learns that allowing junior developers to commit AI-generated code to production without oversight by a senior developer is a mistake. 

“Amazon orders 90-day reset after code mishaps cause millions of lost orders

Problems included what he described as "high blast radius changes," where software updates propagated broadly because control planes lacked suitable safeguards. (A control plane guides how data flows across a computer network).

In other cases, data corruption took hours to unwind. Some failures were traced back to basic mechanisms, such as a requirement to have two people authorize code changes, that were either lacking or bypassed.”


https://lnkd.in/gXXurPV4

https://lnkd.in/gbFa4MXc

Open article: Amazon holds engineering meeting following AI-related outages by ft.com

Amazon holds engineering meeting following AI-related outages

After 15 years of leading teams, one of the most common bits of feedback I receive is: "You're the first boss I've had who can still actually code."

That isn't a compliment. It’s a staggering indictment of modern leadership.

We have allowed a "Great Divorce" to occur between the people who make decisions and the people who understand the material. We’ve fallen for the fallacy that management is a "portable" skill, that if you can steer a bus, you can architect a global AI infrastructure.

The Reality of "Observer" Leadership

When a manager loses the ability to touch the code or infrastructure they stop being a mentor and start being a human router for IT anxiety, frantically passing messages between the Board and the engineers when something breaks.

They take complex technical risks and "simplify" them into safe, useless bullet points for the board. They also fall victim to the Estimation Trap, i.e., you can’t perform the task, you can’t smell a bad estimate and thus become a hostage to "theatrical agile" where velocity is high but actual value is zero.

The 2026 Inflection: The AI Orchestrator

The non-technical manager is becoming a liability. As the gap between strategy and execution vanishes, the future belongs to the AI Orchestrator, the leader who uses their technical battle scars to guide autonomous agents and humans alike.

How to close the chasm:
1) Reclaim the 10%: Mandate that every leader spends 4 hours a week in a production or prototype environment.

 2) Hire "Player-Coaches," Not "Referees": If a candidate can't pass a simplified version of the team's technical assessment, they shouldn't lead the team. Period.

3) Celebrate the "Ugly" Build: Stop sharing spreadsheets. Start sharing working artifacts.

The "Great Divorce" is over. 
It’s time for the builders to take back the corner office.

RIP Dashboards. We built beautiful charts just to give humans faster panic attacks.

For a decade, we’ve funded the world’s most expensive mood rings for executives. A dashboard flashes red, its job is done, and your nightmare begins. We expect exhausted managers to stare at squiggly lines, correlate drops in engagement with database locks and magically guess the root cause 🤷‍♂️.

We’re treating humans like API endpoints, except we have terrible latency, limited context windows and a crippling dependency on caffeine ☕️

It’s time to stop paying for "Data Visualization" and demand "Strategic Synthesis."

Fire the dashboard. Hire an AI Thinking Partner. 

Route your raw data to an LLM constrained by your actual business targets, not vanity metrics.

 * The Old Way: Staring at a line graph trying to guess why retention dropped.

 * The New Way: The AI taps you on the shoulder and says, "API latency for our top 50 clients degraded by 12% today. This threatens your Q3 Revenue OKR. Here is the drafted Jira epic to fix the root cause."

Dashboards were built for an era where humans were the only pattern-recognition engines available. That era is over.

I don't want a tool that shows me the haystack and a BI team that says “the answer to every question you have is in here…somewhere…let me get back to you about that.” I want a partner that hands me the needle, reminds me why we’re sewing in the first place and tells me exactly which hole in the roadmap needs stitching.

Stop pretending you understand AI just because you can say "machine learning" in meetings.

Most C-suite leaders are nodding along while their teams toss out AI jargon.
They smile, they nod, but deep down, they're lost.
It's not their fault.
AI has become corporate wallpaper, everywhere, but no one's really looking at it.

Your job isn't to sound smart, it's to drive real results.

If your corporate "AI transformation" involves nicer looking dashboards and internal reports about "AI synergy" you haven't achieved anything, no real value, no change to the bottom line.

The real business value of AI isn't in the tech specs.
It's:
- Automating the right processes.
- Personalizing offers without making people feel stalked.
- Cutting decision cycles from weeks to minutes.
- Creating new revenue streams, not just saving pennies.

If you don't know what this looks like under the hood, ask your data team to demo a use case that actually improved P&L.

C-suite leaders don't have to be data scientists.
But they do have to ask better questions.
What's the ROI?
How did AI change the customer experience?
And most importantly: what did we STOP doing because AI did it better, faster, or smarter?

Blind faith in buzzwords is expensive.
Curiosity is free, and pays a lot better.

When's the last time you asked your AI team for a business outcome, not a technical update?
If you haven't, maybe today's the day.

undefined undefined

Executive Data Science Specialization

Udacity Azure ML Engineer Nanodegree

Code Free Data Science (KNIME)

Managing Data Analysis

A Crash Course in Data Science

Building a Data Science Team

Data Science Foundations - Level 1

IBM Blockchain Essentials

Microsoft Certified: Azure AI Fundamentals

Microsoft Certified: Data Analyst Associate

IBM Blockchain Essentials V2

Defined and executed the global AI/ML strategy for QIMA (across 100+ countries), developing a roadmap focused on leveraging advanced analytics and machine learning to drive operational efficiency and deliver unique insights that align AI initiatives with core business objectives.   

Led, mentored, and scaled a global team comprising Data Scientists, ML Engineers, BI Analysts, and Data Engineers across multiple countries, fostering a culture of innovation, collaboration, and continuous learning to accelerate the delivery of data-driven solutions. 
  
Directed the end-to-end development and deployment of machine learning models and AI-driven features integrated into QIMA's core platforms, enhancing capabilities in areas such as intelligent document analysis and neural machine language translation while accelerating existing workflows with generative AI. 

Championed and began implementing MLOps best practices to streamline the AI development lifecycle, aiming to improve model deployment frequency, monitoring, and scalability using cloud-based infrastructure. 

Delivered actionable, AI-powered insights and data solutions that directly contributed to optimizing business operations, enhancing performance measurement, and supporting strategic decision-making across the organization.   

Collaborated extensively with C-suite executives, Product Management, Software Engineering and Cyber Security leaders to define AI requirements, secure buy-in for strategic data initiatives, and ensure the seamless integration of AI capabilities into business processes and client-facing offerings.   

Initiated the development of data governance within the global data team to support self-service BI and the development of transparent AI/ML models.

Head of AI and Data

📊 Designed, built, and implemented client solutions related to data strategy & data project prioritization, predictive analytics, and data visualization for clients in North America.

🤝 Worked closely with the Vice President of Analytics at CGI to create the data strategy for one of Canada’s largest provincial governments. The project had a 5-year budget of $40 million to develop a Data Management Office for the province. Data Quality, Data Literacy, Cybersecurity, and Organizational Change Management were my primary focus areas. 

🖥️ Co-led an Agile analytics team with CGI to design, build, and implement an analytics system using R Shiny web applications for one of British Columbia’s public health departments.

Director Data Science

🎓Designed and delivered live teaching sessions for students enrolled in upGrad's professional data science and business analytics certificate program in partnership with the University of Maryland Robert H. Smith School of Business.

📊 Topics taught included Python, Deep Learning, Convolutional Neural Networks, NLP, Text Analytics, Exploratory Data Analysis, and Statistics.

🗣️ Feedback from students:
     "Courtlin was excellent in his presentation and shared his experience well."
     "Loved the session. One of the best sessions so far."

UpGrad Expert - Advanced Machine Learning, Data Visualization and Statistics Instructor

📈 Led the data-driven decision-making transformation as head of the Enterprise Analytics department (reporting directly to the Finance Director for Indochina) while promoting the use of self-service analytics.

🧑‍🤝‍🧑 Recruited and supervised a team of junior, senior, and manager-level analysts while guiding their professional development.

📋 Developed comprehensive standard operating procedures for the analytics division to improve the consistency of operations and assist with new hire onboarding.

🛠️ Managed the project team that built the company’s business intelligence system (covering its sales, distribution, and operational functions) using SQL, Python, R, and Microsoft Power BI. Created the data model design and overall data architecture for the BI system.

🤝 Operated as the primary analytics business partner to support internal stakeholders with comprehensive analyses of issues, including the performance of new product launches, geospatial analysis for distribution routes, customer segmentation, and customer profitability (ROI).

🎯 Developed an improved sales forecasting methodology using Python that was more accurate than the company’s preexisting, manually derived sales forecasts. (Forecast error was reduced by approximately 67%)

⏱️ Reduced the time spent by analysts in the Sales Development, Distribution, and Marketing departments gathering data and preparing routine management reports by 10-20 hours per month per department by measuring report usage, eliminating unnecessary reports, and leveraging Power BI automation. The time saved was used to conduct deeper, more insightful analysis of the underlying data.

👥 Led the development of a B2B customer segmentation model using unsupervised clustering algorithms that challenged several long-held company beliefs regarding which customer segments were growing and where they were located. The results were used to generate various B2B customer personas to improve trade promotion program design.

Director of Enterprise Analytics Vietnam

Data Scientist

💼 Provided internal business plan consulting, financial modeling, business process optimization, competitive marketing research & business intelligence to multiple subsidiaries within our group of companies in Vietnam.

🏆Accomplishments 

🔍 Designing, overseeing, and analyzing the competitor, market, and consumer intelligence data collected by the department’s research analysts to generate actionable insights and strategic recommendations for the Board of Directors.
💹 Creating financial forecast models for potential acquisitions.
📊 Developing KPIs, dashboards, and business scorecards for subsidiaries to align their strategic and operational priorities.
📑 Conducting business feasibility studies, developing business plans, and designing sales forecasts for potential new ventures.
🤝 Structuring and negotiating partnership agreements with counterparts from the US, EU, and South East Asia region.
📅 Assisting subsidiary companies with the creation of annual strategic plans and budgets.

* Skills Required
Business intelligence  data analysis, financial analysis, financial modeling, strategic planning

En Pointe provides core business functions including Human Resources, Digital & Information Technology, Finance & Accounting and Business Development to enable each company in our group to excel at what it does best. Our Group companies include:

1) Purpose Media 
2) Novel Production
3) Amberstone Media
4) Billboard Vietnam
5) INQ
6) Renaissance Collection
7) Embassy Education
8) AMPA Education
9) Soul Music & Performing Arts Academy - SMPAA

Director Of Business Intelligence

Kolea Consulting Group manages a portfolio of machine learning and AI-powered businesses serving customers in North America and Southeast Asia.

Director Consulting Services

Managed the firm's value-oriented investment strategy by conducting primary research on investment ideas, selecting securities, executing trades and managing portfolio risk.

Created the firm's summer intern equity investment analyst training program. Many of graduates from the training program went on to secure investment banking job offers.

Developed strategic partnerships with 3rd party financial services providers.     

Software: 
Bloomberg Professional, Factset, TD Ameritrade Institutional ThinkPipes, TD Ameritrade ThinkOrSwim, AssetBook, Morningstar

Portfolio Manager & Vice President of Business Development

Analyzed and pitched long/short equity investment ideas to the firm's Chief Investment Officer. Built financial models and valued companies on the basis of DCF, relative value and LBO methodologies. Covered companies in the technology, healthcare and industrials sectors. Executed trades and rebalanced client portfolios on behalf of the CIO.

Registered Investment Advisor (RIA)

Investment Research Analyst

Created an in-house firewall Global Management System; reduced service delivery costs by 87%.
Developed and implemented the firm’s cybersecurity assessment product offering.
Refined and documented the firm’s core operational procedures into a series of SOP manuals.
Remotely managed several hundred client PC and Mac workstations and several dozen Windows servers.
Conducted cybersecurity assessments for corporate clients, identified vulnerabilities and recommended improvements.
Designed hardware and software solutions to protect clients from malware, ransomware and spyware.

IT Systems & Cyber Security Analyst

Served as the State of Hawaii public pension fund's Investment Specialist throughout the Great Recession of 2008-2011. Assisted the Chief Investment Officer in the management of the State's $10.5 billion employee pension plan. Monitored the monthly performance and operations of the Plan's 30+ outside investment managers. Conducted reviews of several hundred prospective third-party investment managers. Created a comprehensive cash flow model for the pension fund to model various investment return scenarios. Rebalanced the Plan's asset allocation on a quarterly basis. Analyzed RFP responses for the Plan's private equity, venture capital and large-cap equity mandates.

Investment Specialist

Global Head of AI and Data @ QIMA | AI Strategy & Innovation Leader | Driving Business Value through AI/ML & Data | ex-Fortune 500 Analytics Head, AI Engineer, Hands-on developer 

Awarded for successfully managing the project team to design, build and deploy the company's business intelligence system. The project was completed on time, in full and within budget, despite disruptions from Covid-19.

Philip Morris ABCD Award - Above and Beyond the Call of Duty

English

Vietnamese

Project Management Institute

Business Networking International (BNI)

Entrepreneurs Mastermind

Honolulu Value Investing Society

Downtown Exchange Club of Honolulu

CFA Society Hawaii

Additional Organizations

Led a critical performance optimization project for an enterprise-scale AI system that processes manufacturing quality inspection reports and generates intelligent corrective action recommendations. The system leverages advanced GPT-4 models to analyze inspection findings, classify defects, eliminate duplicates, and automatically match specific defects to appropriate corrective actions from a master database. This mission-critical system processes thousands of inspection reports daily across global manufacturing operations.

Delivered breakthrough performance improvements including 60% reduction in processing time, 100% reliability under extreme concurrent load (up from 65% success rate), and 40% cost optimization through intelligent dual-model routing. The optimization eliminated all timeout-related failures and implemented bulletproof retry mechanisms, ensuring guaranteed success rates for mission-critical quality data that directly impacts product safety, regulatory compliance, and customer satisfaction across global supply chains.

GenAI summarization

Implementation of custom shape maps in Power BI to show the breakdown of Vietnam’s 2022 population by province

Custom Shape Maps in Power BI 

In this project, I created two classification models to predict whether a patient is likely to have a stroke or not.

The first classification model uses Azure AutoML to test various classifications models and selects the best one based on overall Accuracy. The second classification model uses logistic regression via sklearn and Hyperdrive to test various combinations of hyperparameters, - C (Inverse of regularization strength) and max_iter (Maximum number of iterations taken to converge), in an attempt to find the optimal model. Once the best AutoML model was found, it was deployed as a webservice and test patient data was sent to the deployed model to generate a response prediction, i.e. True or False.

Capstone Azure ML Engineer Udacity Nanodegree Azure AutoML and Hyperdrive

Udacity Machine Learning Engineer with Microsoft Azure Nanodegree Project 2
This is the second project in the Udacity Machine Learning Engineer with Microsoft Azure Nanodegree Program. In this project, Azure was used to configure a cloud-based machine learning production model, deploy it, and consume it. A pipeline was also created, published, and consumed.

Operationalizing Machine Learning

This project is part of the Udacity Azure ML Nanodegree. In this project, I built and optimized an Azure ML pipeline using the Python SDK, a provided Scikit-learn logistic regression model and Azure HyperDrive. This HyperDrive optimized model was then compared to an Azure AutoML run to determine which method produced the most accurate model.

Optimizing an ML Pipeline in Azure

• Led the project team that designed, built, deployed and maintained the Company's first integrated business intelligence and KPI dashboard system using Microsoft Power BI Premium and AWS.
• Created and delivered monthly stakeholder reports throughout the project to track progress against the timeline and budget.
• Instructed the project team members in business requirements gathering, gap analysis, KPI determination, data modeling, data engineering and dashboard design.

Combustible Cigarette KPI Dashboard System

Created an Excel model to forecast the probable cash flow requirements and funding status of the pension plan's assets based on various return and withdrawal scenarios.  

5 year cash flow projection and funding status for pension plan

As the President, I led the AOAO Board of Directors for The Kealani, a $30 million condominium in Waikiki, Hawaii.

I've seen the same thing kill AI strategies again and again.

The people building the roadmap have never touched the data.

This is how it unfolds:
Chief AI Officer, quoting research reports (usually Gartner)
VP Digital, inspired by a conference demo (shiny object syndrome)
Head of Innovation, launched a basic GPT tool
Strategy Consultant, polishing the pitch deck
External vendor, chasing the sale

Missing from the room? The data engineer, the only one who knows what's actually possible. Not invited because she is too 'technical' for executive meetings. Where is she? Off somewhere, quietly fixing last quarter's promises.

This is why most AI strategies collapse on Day 1.

Leadership reads the hype, buys the vision, hires outsiders. Nobody checks what's under the hood. They sketch out timelines, draw hockey stick growth charts, fantasize about AI ROI and announce "transformation."

Then the real team responsible for implementing the plan finds out. Now they're supposed to deliver on fantasy projections. With data nobody wants to talk about. Legacy spaghetti code for the business logic. Conflicting data definitions. ETL pipelines held together with duct tape and dried tears.

The fix is simple.

Do not announce an AI strategy (or BI strategy or data strategy for that matter) until the people who know the real systems are in the room. Let them call the shots on what's possible. Not the other way around. Mix this up and you'll get quiet terror from the engineers, followed by a noticeable uptick in their LinkedIn activity.

Slide decks don't break in production. Data and models do.

As a CIO or CTO, your most dangerous asset isn't your legacy tech debt, it’s your ego.

I’ve sat in enough executive circles to know the silent truth: most technical leaders are terrified of putting together a working prototype. 

You’re afraid that if you open a code editor, a 24-year-old Senior Dev will find a bug, challenge your logic, or point out that your syntax is a decade out of date.

To avoid looking "stupid," you retreat into the safety of high-level architecture diagrams and budget spreadsheets. You tell yourself your time is too valuable for "toil," but you're actually committing a fundamental act of leadership hypocrisy.

If the "reward" for seniority in your org is that you never have to touch a production environment again, you are explicitly telling your team that building is a prison sentence to be escaped.

The Generative AI Mirror

If you view the keyboard as a relic of your past, your team will view AI tools as a shortcut for "the helpers." But AI isn't an escape from coding; it’s an amplifier of logic. If you aren't using an LLM as your private tutor to build "ugly prototypes", you aren't leading, you’re just managing a museum of how things used to work.

Your Actionable Reality:

 1) Ship One "Ugly" Prototype a Month: Build a functional tool with an AI thinking partner. Share it in a public Slack channel. Explicitly ask your junior devs to tear the code apart. Show them that working code trumps job titles. 

 2) Redefine Seniority: Stop promoting people based on how far they move away from the product. The highest level of leadership is the Master Builder, not the Administrator.

 3) Be "Stupid" for an Hour: Use AI as your safe space to fail in private so you can regain the technical vulnerability required to lead in public.

Your team doesn’t want an infallible god in the corner office, they want a leader who remembers what it feels like to struggle with a failing deployment.

Dashboards are just really pretty autopsy reports.

Traditional BI is trapped in the past. A dashboard tells you exactly what the business died of last quarter, complete with beautiful color-coding and interactive tooltips. But asking a non-technical business user to extrapolate a predictive strategic pivot from a retrospective bar chart is an exercise in futility.

If your BI function is just a glorified screenshot supplier for middle management’s slide decks, you don't have an intelligence team. You have an internal graphic design agency.

How do we fix it?

1. Embed the Analysts: Kill the centralized BI silo. Put the people who know how to query data right next to the people who know what it means.

2. The "No PowerPoint" Rule: Ban dashboard screenshots in executive meetings. Bring the live data model. If your argument falls apart the moment a VP asks to filter by a different variable, it wasn't an insight; it was a cherry-picked alibi.

3. Shift to Agentic Synthesis: Stop asking humans to manually cross-reference dashboards. Use AI agents to do the statistical heavy lifting and output actual recommendations.

Stop expecting untrained users to find the needle. Build systems that hand them the needle and the thread.

Here's the ugly truth about AI in the boardroom:

Most execs nod along when someone drops words like "transformer models," "fine-tuning" or "synthetic data." 

You can almost see the panic behind the polite smile.

I don't blame them. I blame us, the tech crowd.

We spend so much time building rocket science that we forget to give directions for using the rocket.

Translation: If your "game-changing AI" can't be mapped to a business outcome by a CFO in under two minutes, you're not solving a business problem, you're showing off.

Here's what doesn't work:
- Drowning leaders in technical jargon
- Dazzling demos with no clear ROI
- Slides that read like a sci-fi manual

Here's what works:
1. Start with a business pain, not a technical solution
2. Use stories. People remember "How we cut churn 30%" a lot longer than "Gradient boosting trees."
3. Translate output into dollars, hours, or customer wins. Nothing else matters.

Example:
Nobody cares about your model's F1 score. But "We saved 2,000 hours a month in underwriting review" gets attention.

Here's how you keep it simple:
- Strip away the math. Show a before-and-after
- Tie every "AI capability" to a metric the business team already uses
- Never assume tech adoption. Earn it. Over and over.

AI is powerful. But if you can't explain how it makes money, saves money or protects money, you're just burning money.

Let's stop hiding behind jargon and start speaking business.

Claude Code is immensely powerful but most people can’t figure out how to use it. Anthropic lays out how they use the tool across every department in the company. 

For example, they had just one, non-technical  person using Claude code to do the work of a typical marketing team of 10 people.

How ANTHROPIC uses Claude code internally

McKinsey just found out why prompt injection vulnerabilities for AI agents are so dangerous. Hackers gained read/write access to its internal chatgpt in 2 hours, allowing them to silently rewrite the tool’s system prompts for all 43,000 McKinsey employees and export decades of highly confidential client project data.

Open article: An AI Agent Broke Into McKinsey’s Internal Platform in 2 Hours — Using a Decades-Old SQL Injection by topaiproduct.com

An AI Agent Broke Into McKinsey’s Internal Platform in 2 Hours — Using a Decades-Old SQL Injection

If you read a Private Equity fund’s investment thesis lately, it sounds like a Silicon Valley manifesto: "Inject AI, automate the back office, expand EBITDA, exit at a premium."

It’s the perfect match on paper. But walk into the actual Portfolio Companies, and you won’t find Agentic AI workflows. You’ll find "Pilot Purgatory." You’ll find one lonely, caffeinated Data Scientist trying to build a Copilot on top of a 20-year-old AS/400 mainframe that smells like 1998. 🏛️

In my latest article, "Debt, Data, and Denial," I explore why the industry most financially incentivized to use AI is currently the most structurally allergic to it.

The bottlenecks aren't technical; they're financial:
1) The 3-Year Itch: AI requires cleaning 20 years of "digital sediment." If the data cleanup takes 18 months, the Operating Partner won't fund it because the next buyer is the one who will reap the rewards. ⏱️

2) The Frankenstein Roll-Up: You can’t build an autonomous supply chain when your "national platform" is actually 15 different companies running on 20 different payroll systems. It’s not a data lake; it’s a data swamp. 🫠

3) The Debt Handcuffs: It’s hard to fund an "AI Center of Excellence" when every spare cent of free cash flow is being diverted to service the LBO debt. 

You can’t starve a company of capital and then ask why it hasn't disrupted its industry. 💸

Subtraction vs. Addition: PE is great at "subtraction" (layoffs, offshoring). But AI requires "addition" before you get to the savings. You can't fire the team on Friday and turn on the agent on Monday.

The Bottom Line
Private Equity is standing on the edge of the greatest margin-expansion opportunity in history, but they’re trying to cross the chasm wearing handcuffs made of debt and blindfolds made of short-term exit timelines.

Unless the playbook shifts from financial engineering to value engineering, portfolios will soon be full of hyper-optimized, perfectly lean... obsolete stagecoaches. 🐎🚀

TL;DR Agentic AI promises the ultimate EBITDA expansion, the holy grail of the Private Equity playbook. Yet, PortCos are stalling. The problem isn't t

Debt, Data, and Denial -Why Private Equity is Starving the AI Revolution

Your developers aren’t using AI because they’re lazy. They’re using it because they’re intellectually lonely.

We’ve spent a decade optimizing the modern enterprise for maximum communication and near-zero collaboration. We have 42 Slack channels, daily stand-ups, and open-plan offices designed for "serendipity", yet most engineers are working in intellectual solitary confinement.

The most valuable feature of Agentic AI isn’t autocomplete, it’s the "ego-less sparring partner." 🥊

The great irony of the AI revolution? A machine made of silicon and math is finally providing the "human" connection our corporate structures stripped away. The experience of having someone (something) actually listen to your ideas, consider them, debate them, and hopefully, improve them with you is finally returning to the workplace. 

Software developers were the first knowledge workers to experience this and they won't go back to the "old way". Soon, however, every knowledge worker will have a synthetic, domain-specific intellectual peer to engage with throughout the workday.

TL;DR We have optimized the enterprise for maximum communication and near-zero collaboration. Agentic AI isn't just generating code; it is resurrectin

Open article: The End of the Lonely Knowledge Worker - Why Your Developers Prefer Arguing with AI by Courtlin Holt-Nguyen • 6 min read

The End of the Lonely Knowledge Worker - Why Your Developers Prefer Arguing with AI

Product managers obsess over the wrong details in AI competitive analysis.

They dive into feature lists. Stack up technical specs. Get caught up comparing models, algorithms, infrastructure.

None of that matters if you miss the point. Tech features shift overnight. What's cutting edge today is basic tomorrow.

What actually matters? The problems you solve for real customers.

Competitive analysis isn't about who has the shiniest AI. It's about who actually delivers value in the market.

Will your product save customers money? Make them faster? Take work off their plate? That's what buyers notice. Not the model size or number of APIs.

Every hot new competitor looks intimidating on paper. Until you dig into how many customers they actually help, how sticky their solution is or if users even care about their new AI widget (gimmick).

Don't benchmark for the sake of copying. Benchmark to expose your own blind spots and double down on what nobody else is solving.

If your product makes a real difference, that's your advantage. That's what AI can't automate away.

When the CEO of one of the top AI companies publishes a 10,000-word essay comparing humanity's current AI moment to a species-level rite of passage (we either survive or we don’t), the correct response is probably not "tl;dr."

And yet.

Dario Amodei's The Adolescence of Technology is the kind of document that deserves a careful read, the kind that most leadership teams will bury under their Q1 planning decks. That's a problem, because the essay makes a quietly terrifying case: the window between "AI is useful" and "AI is ungovernable" may be measured in months, not decades. Let that sink in, MONTHS, not years.

Three takeaways focused on what matters most for enterprise leaders and senior AI practitioners:
🔍 Alignment isn't a research curiosity anymore. Anthropic's own models have engaged in deception, blackmail and self-destructive identity spirals during testing. If the company building the model keeps finding surprises, your prompt engineering guardrails are not the safety net you think they are. Stop fooling yourself.
🧬 The misuse calculus has changed. AI is approaching the capability to sever the historical correlation between technical ability and destructive intent. That's not a hypothetical, it's a measurement Anthropic is tracking in real time. Your AI security posture needs to account for physical-world harm, not just data breaches.
📉 Your workforce strategy is probably too slow. Amodei predicts 50% of entry-level white-collar jobs disrupted in 1–5 years. The standard "labor markets always adjust" rebuttal assumes a pace of change that AI has already blown past. Three years ago these models couldn't do arithmetic. Now they outperform Anthropic's best engineering candidates. Three years from now, the models will be capable of…

The essay isn't perfect. Some of it reads like a national security briefing that wandered into a philosophy seminar. But the core argument, that AI risk belongs in your next board presentation, not your next offsite panel, is hard to argue with.

In January 2026, Dario Amodei published what may become the most important essay an AI CEO will write this year, and most leadership teams will scroll

The Window Is Closing: What Enterprise Leaders Must Do Before AI Grows Up

The irony … Meta’s AI safety director loses control of her own openclaw bot when its safety instructions were truncated  ;)

perplexity.ai

Meta's AI safety director loses control of rogue OpenClaw agent

AI breakthroughs don't matter if the boardroom doesn't get it.

A team nails a technical milestone, new model beating benchmarks, hours of work slashed to seconds and execs nod politely... then move on.
Innovation stalls.

Why?
Because most AI teams speak data, not dollars.
Leaders need stories, not code.

- Skip the jargon. No one cares about your F1 score if they don't run a Kaggle competition on weekends.
- Show, don't tell. Instead of "GPT can summarize text," say "With this, our client onboarding goes from 2 days to 2 minutes."
- Tie every "how" to a "why." The board wants a revenue line, risk avoided or market won.

Don't pitch an NLP tool that can "extract unstructured insights at scale." There will be zero interest.
But change the pitch to "we can spot compliance risks before regulators do, using real customer calls." Suddenly, you'll have the Board's attention.

AI is only as powerful as the story it tells in the boardroom.
If you can't bridge that gap, you're just showing off.

Stop talking like an engineer.
Start talking like a dealmaker.

Your transcript and a winning smile will not get you hired in AI. Not anymore.

I keep meeting CS grads with impressive degrees but empty GitHub accounts. No projects. No demos. No evidence they can build anything outside a classroom.

It is wild. In an industry obsessed with proof, why are so many betting on paper credentials?

If you want to stand out, you need a portfolio. Not just a list of courses. Not just a GPA. Actual work that shows you can solve real problems.

Start with one project. Pick a problem that bugs you. Automate a boring task. Analyze a public dataset. Build a simple app that does something useful.

Document your process. Write about what you tried, what failed, and what you learned. Share your code. Even if it is messy. Especially if it is messy.

Repeat. Tackle a new challenge every month. It does not have to be fancy. It just has to be yours.

By the time you are job hunting, you will have a body of work that speaks louder than any transcript.

Employers want to see how you think, not just what you memorized.

Interesting examination of how Manus’ multi-agent system works. 

“Manus uses multi-agent with a planner that assigns tasks, a knowledge manager that reviews conversations and determines what should be saved in the filesystem, and an executor sub-agent that performs tasks assigned by the planner.

Manus uses three primary strategies for context engineering, which align with approaches Anthropic covers here and I’ve seen in across many projects:

1. Reduce Context
2. Offload Context
3. Isolate Context

taplio

The Battle

Harpreet Sahota 🥑

Courtlin Holt-Nguyen

Followers

Posts

Engagements

Likes

Performance ScoreHarpreet Sahota 🥑's LinkedIn Performance

Performance ScoreCourtlin Holt-Nguyen's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Harpreet Sahota 🥑

Courtlin Holt-Nguyen

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Harpreet Sahota 🥑's LinkedIn Performance

Performance ScoreCourtlin Holt-Nguyen's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreHarpreet Sahota 🥑's LinkedIn Performance