William Szamosszegi

MyWorker.ai

Pioneering “AI Sales Team as a Service” - so founders can scale without limits.
1. Category Creator: Built the world’s first AI Sales Team as a Service, letting ambitious founders and teams instantly deploy a full digital SDR squad - no hiring, no burnout, no time wasted on cold outreach.

2. Unblocks Growth: MyWorker fills your calendar with qualified sales meetings while you stay focused on product, fundraising, or scaling - solving the #1 pain point for startups and SMBs who can’t hire (or don’t want to manage) a traditional sales team.

3. AI Voice in Action: Launched “Alex,” our always-on AI sales rep, to personally call, qualify, and book meetings for clients. Alex works 24/7, never forgets a follow-up.

4. Proven Impact: Helped early customers double lead-to-demo conversion rates, cut founder “busywork” by up to 90%, and accelerate pipeline growth from week one.

5. Mission: Redefining how founders grow - turning outbound sales from a grind into a growth engine powered by AI.

If you’re building and want to focus on your vision, not just your inbox, let MyWorker book your next best customers - while you sleep.

CEO & Founder

Sazmining

President & Chairman of the Board

Making Bitcoin mining the easiest, smartest way to build your stack - sustainably and securely.
1. Democratizing Bitcoin Mining: Transformed mining from a technical challenge into a seamless service. Customers can sign up to become a Bitcoin Miner in minutes - it's easier than buying on an exchange.

2. BTC Dollar Cost Averaging 'on steroids': Top-Rated Bitcoin accumulation through mining, not just buying. Get “wild sats” direct from the source, at lower cost and with zero exchange hassle. 

3. Global Green Operations: Led Sazmining to launch and operate renewable-powered mining facilities in the US, Paraguay (100% hydropower), and beyond. Committed to making Bitcoin mining accessible and environmentally responsible.

4. Education & Trust: Hosted the Sazmining Podcast and spoke at leading Bitcoin events - educating thousands on sustainable mining, DCA, and Bitcoin innovation.

5. Media & Recognition: Featured in Yahoo Finance, MarketWatch, Business Insider, and major industry press. Recognized as an industry leader at the intersection of Bitcoin and clean energy.

Why it matters: Sazmining empowers everyday people to build wealth sustainably - proving that the future of Bitcoin is open to all, not just the few.

TurnUh LLC

Mobile app that shows events happening on your college campus. Integrated with event details, live comments, and up-vote/down-vote system.

Founder and CEO

BHG Financial

First-hand exposure to high-level operations at a large company while working under the Chief Business Development Officer, Tyler Crawford. Managed scheduling, follow ups, note-taking and side projects in helping BHG's decision-makers move the company forward.

Business Development Intern

IRIS.TV

In working with the accounts management team at Iris, I became familiar with online tools like Slack, Salesforce and Asana from my immersion in the fast-paced startup culture. I extracted client data from RStudio, MySQL, Iris Vision and other data mining sources and then complied that data in excel. This included prescriptive analytics, performance, views, retention, experience, lift, bounce rate, quarterly analysis, month over month analysis, and other variables that varied by client. My reports also included graphical representations in the form of heat-maps, and charts which aided in overall comprehension of the data. After, I would meet with one of the account executives, Mantonie, and analyze the data looking for trends or possible issues to help Iris.TV provide clients with in-depth consultation.

Summer Analyst Intern

Tired of juggling 7+ sales tools, endless CSVs, and inbox-warm hacks - yet sales pipeline still flat?
Feel like every new AI promise adds more duct-tape, not more demos?

I’ve been there. So I built MyWorker.ai - the first “AI Sales Team-as-a-Service” that handles prospect research, personalization, deliverability, and follow-ups while you sleep.

🚨 The headaches we remove
1. Messy tool stack → one AI Sales engine that works.

2. Low conversion → opens leap from low-20 % to > 55 % in two weeks. And replies 2x on average

3. Wasted Time → founders win back 80 % of prospecting time to focus on closing.

👉 Who’s a fit
1. B2B SaaS Founders, doing $1–20 MM ARR, < 50 teammates.

2. Prefer calendar full of customers over “another dashboard.”

3. Ready to replace busywork, not just improve it.

📈 Proof in numbers
1. Clients moving off Apollo + Instantly stacks see 2–3× reply rates month one.

2. 57 % average open rate across 210 k emails (same copy, just smarter delivery).

3. Spotlighted in Yahoo Finance, Business Insider, MarketWatch for category-creating growth.

🛠️ How it works
1. 15-min kickoff - import ICP & brand tone.

2. Autopilot ON - AI starts working to book demos next business day.

3. Review mode (optional) - you approve, AI learns, then goes fully hands-off.

About me
• Scaled Sazmining from bootstrap to renewable Bitcoin mining on 3 continents (USA, Paraguay, Norway)
• Forbes Fellow, SXSW Digital Entrepreneurship finalist.
• Co-founded Winter Tennis Special Olympics for youth; obsessed with tech that compounds impact.

🔹Let’s connect if you’re:

1. A time-strapped SaaS founder who wants customer pipeline, not plug-ins.

2. An investor scouting the next AI revenue infra layer.

3. A builder who loves moving fast and measuring everything.

The bold shape the future. Will you wait to see what happens or be the one building it?

Ping me if you’re a Founder ready to ditch grunt work and grow.

CEO & Founder & Secretary

Data processing that fosters innovation in Blockchain Technology, Artificial Intelligence, CGI Rendering, and Research. Our computing power supports the next generation of technological development — Web 3.0.

One mistake cost me millions of views - and it changed me permanently.

During Tech Week, we lost audio for massive interviews:
Corporate Bro, Apollo’s COO, and a few others.

Gone.
Irrecoverable.

It crushed me.
But it forced something I had been avoiding:

I stopped outsourcing my technical competence.

I learned cameras, audio, lighting, backup systems, production workflows.
I rebuilt my entire setup from scratch.

The pain became an upgrade.

Sometimes something breaks so you stop relying on people who can’t move at your speed.

And you come out 10x more dangerous.

Built the #1 Sales Tool in the World | MyWorker.ai, CEO/Founder | Sazmining, President/Founder

last friday reminded me why showing up matters.

I went to an event expecting to just catch up with a few friends.
instead, it turned into one of those nights that quietly resets your energy as a founder.

I saw people I haven’t crossed paths with in months.
I met people I’ve respected from a distance.
I had the chance to speak with Andrew Yang.
and the conversations went way deeper than the typical “networking” vibe.

when you’re building, it’s easy to disappear into the grind.
to get buried under the pressure, the timelines, the constant motion.
you forget that momentum often comes from the one room you almost
skipped…
the one conversation you weren’t expecting…
the one person who reframes something you’ve been stuck on for months.

what hit me that night was simple:
the people you surround yourself with shape how far you can see.
and founders forget that when we’re carrying the weight alone.

execution matters
consistency matters
but so does stepping into the world
meeting people who expand your thinking
and putting yourself in places where luck can actually find you.

I left the event more grounded, more inspired, and more committed to the long game.
sometimes all you need is one good room and the right people in it.

li_at

JSESSIONID

Some pressure builds slowly. Some hits all at once.

I’ve been holding everything together for months - building multiple companies in public, streaming for hours a day, raising capital, debugging product, onboarding new systems, jumping between 9 different channels.

Yesterday it hit me how easy it is to forget that I’m still just one person.

The weight sneaks up quietly.
Then suddenly it’s in your chest.

That mix of purpose, exhaustion, momentum, fear.

People see the output - the streams, the traction, the optimism.
They don’t see the part where you’re trying to keep the entire structure standing while making sure nobody feels the cracks you’re carrying.

I’m learning that the emotional weight isn’t the enemy.
It’s the signal that you’re on the edge of becoming someone new.

Building in public exposes everything.
But it also forces you to grow faster than you thought you could.

Building in public forces a different kind of growth. 

When you build privately, you can hide your confusion.
Your mistakes.
Your fear.

When you build publicly, everything is visible. the decisions, the pivots, the broken pieces.

At first, it’s uncomfortable.

Eventually, it becomes a superpower.

You stop performing.
You start refining.

The market learns you.
You learn the market.
Trust compounds.

Public building is pressure, but it's also clarity.

AI doesn't replace people. It replaces the version of you that refuses to adapt. 

The founders who win in 2025 won’t be the ones with all the answers.

They’ll be the ones who learn faster than the environment shifts.

Leverage isn’t optional anymore.
It’s the cost of staying relevant.

Every week something inside MyWorker resets my perspective. 

This week it was seeing how quickly someone can go from “I don’t get this” to “I can build with this.”

I’ve seen it with my assistant.
With customers.
Inside my team.

The barrier to entry is collapsing.
The barrier to adaptation is rising.

The winners of the next decade won’t be the most technical.
They’ll be the ones who can translate intelligence into motion.

Distribution isn't content. It's culture. 

Most founders think “posting more” is distribution.

It’s not.
That’s just broadcasting.

Real distribution is cultural fluency.

Understanding Discord tribes.
Streamer ecosystems.
Creator language.
Internet micro-dialects.

The internet rewards the people who can code-switch between digital cultures.

The future belongs to founders who can move through these spaces without breaking flow.

Everyone is still underestimating what's coming. 

AI doesn’t scale like software.
It scales like intelligence.

Compute gets cheaper → models get smarter → models improve how we use compute → the cycle accelerates again.

That flywheel is moving faster than society can absorb.

The moment it clicked for me was onboarding my assistant.
She went from overwhelmed → extremely capable in a matter of hours.

The baseline skill level in society is unbelievably low.
Once AI becomes even slightly more intuitive, output will explode for the people who lean in - and collapse for everyone else.

We’re about to see a gap form that nobody is prepared for.

And it won’t be about money or access.
It’ll be about willingness to adapt.

People see momentum. They don't see the weight. 

Right now I’m building multiple companies in public.
Raising capital.
Streaming daily.
Fixing product.
Onboarding teams.
Running events.
Rebuilding distribution loops.
Navigating Discord.
Trying to be everywhere at once.

And at the same time, I’m holding the emotional energy for everyone around me - optimism, stability, certainty.

The truth is: some days it feels like everything could collapse.
Other days it feels like I’m inches from the biggest breakthrough of my life.

That tension is the actual founder journey.
Not the highlight reel.

Founders underestimate how chaotic real distribution is. 

The deeper I go into Discord and creator culture, the more it feels like the early Bitcoin mining days - messy, unpredictable, tribal.

No rules.
No clean frameworks.
Just raw chaos.

Yesterday I spent hours hopping between servers, absorbing the culture in real time, talking to people who basically live online.

And it hit me:

The future belongs to founders who can navigate chaos and culture at the same time. 

You can’t outsource this.
You can’t “hire a community manager” and hope they understand the culture better than you.

Distribution is becoming a cultural skillset, not a marketing function.

The founders who win will be the ones who learn how to move inside the chaos without losing themselves.

Klaviyo just deplatformed Sazmining

not for wrongdoing.
not for risk.
not for customer complaints.

they shut us down because we help people acquire bitcoin the same way Satoshi did.

through mining.
through energy.
through proof of work.
through a system that doesn’t require permission from anyone.

and that’s the part no one wants to say out loud.

platforms aren’t banning “crypto.”
they’re banning the one thing that gives people the ability to opt out.

bitcoin mining is the most transparent, verifiable, regulated, cleanly-audited way to acquire bitcoin on earth.
it’s the opposite of the speculation they claim to be protecting against.

but when you threaten the old model
when you give people a way to participate without intermediaries
when you make sovereignty accessible
the system treats you like a problem

it’s easier for them to lump bitcoin into the same category as whatever token launched last week than to admit what mining actually represents.

permissionless access.
self-custody.
energy-backed independence.
a level playing field.

this wasn’t a compliance decision.
it was a censorship moment disguised as policy.

but here’s the thing:
bitcoin was built for moments like this.

I’m not discouraged.
I’m not slowing down.
and I’m definitely not asking for permission.

Sazmining will keep building the infrastructure that gives people the same freedom Satoshi had.
the same freedom every miner has.
the freedom to participate without needing anyone’s approval.

they can shut off an account.
they can’t shut off the network.

they can keep their gatekeeping.
we’ll keep our hash rate.

We’re migrating to a provider that actually understands bitcoin.

If you know a company aligned with that mission, tag them below.
and if you’re against this kind of censorship, share this so the story gets louder.

Rasa Hero. NLP/NLU, Chatbots, Voice, Conversational UI/UX, CX Designer, Developer, Ubiquitous User Interfaces. 

https://cobusgreyling.medium.com
https://www.eliza.community
https://www.cobusgreyling.me

CX requires four ingredients:

1. Continuity: Vertically in one Medium & Horizontally across Mediums.
2. Intelligent Directed Dialog. 
3. Contextual Awareness. 
4. Intent Discovery

Cobus Greyling holds 4 tertiary qualifications, including a Post-Graduate diploma from the University of Pretoria in GIS. 

Over a period of more than 20 years he has visited 19 countries, and has worked in 5 countries, with 8 GSM networks, and more than 18 companies. Being involved with self-service in 14 languages on various platforms relating to IVR (inbound, outbound, ASR and Speech Synthesis), Voice Biometrics, Conversational Interfaces (Amazon Echo, Google Home, Chatbots), Cognitive Computing, Linguistic Analysis and Ambient Computing. 

Cobus has experience in the areas of training, development, project management, and executive overviews and presentations. 

Cobus also has experience in GIS (Geographic Information Systems), understanding spatial analysis etc.

Countries worked in:
South Africa (Various companies), Lesotho (Vodacom), Tanzania (Vodacom), DRC Congo (Vodacom), Nigeria (VMobile, Celtel), Finland (Tecnotree)

Countries visited:
United States, England, Congo Brazzaville, Namibia, Botswana, Zambia, Zimbabwe, Egypt, France, Netherlands, Switzerland, Belgium, Germany and Italy.

Cobus Greyling

From autonomous agents reshaping support economics to $100M ARR milestones, the customer service AI market is scaling fast. CB Insights identified the current leaders and their challengers based on their market share...

Chief Evangelist: AI Agents & Agentic Workflows

Microsoft analysed > 37 mil conversations on Microsoft Copilot to understand how it is used...

Unlike prior analyses of AI usage, they focus not just on what people do with AI, but on how and when they do it. 

Surprise-surprise, however we use AI depends on context and device type...the medium matters...

On mobile, health is the dominant topic, which is consistent across every
hour and every month they observed—with users seeking not just information but also advice. 

On desktop, the pattern is strikingly different: work and technology dominate during business hours, with “Work and Career” overtaking “Technology” as the top topic precisely between 8 a.m. and 5
p.m. 

These differences extend to temporal rhythms: programming queries spike on weekdays while gaming rises on weekends, philosophical questions climb during late-night hours, and relationship conversations surge on Valentine’s Day. 

These patterns suggest that users have rapidly integrated AI into the full texture of their lives, as a work aid at their desks and a companion on their phones.

FlowMind Is An Automatic Workflow Generator
The FlowMind research was conducted by JPMorganChase to use LLMs to overcome the rigidity of RPA. FlowMind leverages a prompt template to ground LLM reasoning via APIs.

Data integrity and confidentiality is maintained while generating automated workflows. The effectiveness of FlowMind is demonstrated with success in workflow generation, user interaction & feedback.

FlowMind reminds much of Autonomous Agents as we have come to know them. In particular LangChain Agents and LlamaIndex Agentic RAG comes to mind.

FlowMind & Agents...

In both of these implementations the agent creates a sequence or chain of events (analogous to automatic robotic process creation) which is executed.

The only difference in this regards is that FlowMind creates the process or flow as a Python function. Which it then in turn executes to deliver the answer.

Prompt Recipe & Prompt Template

At the heart of agents as we have come to know them, is a prompt template.

FlowMind makes use of something called a Prompt Recipe, which is shown below. The prompt recipe as JPMorganChase calls it, is in essence a template.

CB Insights mined their Business Graph to map the 135+ startups across 17 markets forming their new AI agent tech stack. 

I found the graphic showing the different logos, categories and sub-categories most insightful.

CBINSIGHTS - The AI agent tech stack

Why aren't more people talking about the massive advances in 𝘁𝗼𝗼𝗹 𝗳𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝗮𝗹𝗶𝘁𝘆 and 𝗮𝗴𝗲𝗻𝘁𝗶𝗰 𝗰𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 for LLMs?

LLMs are rapidly gaining true 𝗮𝗴𝗲𝗻𝗰𝘆—the ability to autonomously plan, reason over long horizons, reliably call tools (in parallel on series/chained) and execute complex multi-step tasks end-to-end.

Just this week, OpenAI rolled out 𝗚𝗣𝗧-𝟱.𝟮, with huge leaps in 𝗮𝗴𝗲𝗻𝘁𝗶𝗰 𝘁𝗼𝗼𝗹-𝗰𝗮𝗹𝗹𝗶𝗻𝗴, long-context reasoning and seamless integration of domain-specific tools via the SDK. 

This tighter coupling between model and runtime reduces latency, cuts errors in workflows, and makes building powerful agents far easier. (BTW, I created a Python notebook with his functionality to prototype it...impressive to say the least.)

Meanwhile, xAI pushed the envelope with Grok 4.1 Fast and its Agent Tools API—server-side orchestration of tools like web search, code execution, and more, letting the model act with real autonomy without developers managing every detail.

To me, "agentic" means the system has genuine freedom to decide, sequence, and iterate using tools—turning LLMs from smart responders into proactive doers.

Developers define custom tools once, and the model intelligently plans and executes around them.

Yet amid all the benchmark wars and flashy demos, these foundational shifts in tool management  and agentic behaviour feel underrated. 

They're the quiet revolution enabling reliable, production-grade AI agents.

(In my latest article on Medium, there's a full working Python example you can copy-paste into Colab to try this yourself.)

Salesforce 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 (Weiran Yao) released an 𝘦𝘯𝘵𝘦𝘳𝘱𝘳𝘪𝘴𝘦 𝘥𝘦𝘦𝘱 𝘳𝘦𝘴𝘦𝘢𝘳𝘤𝘩 (EDR) which is a multi-agent enterprise-grade deep research framework.

It consists of a feature rich GUI, and an underlying agentic workflow. With a master planning agent, and multiple worker sub-agents. Human-In-The-Loop, visualisation, reflection and MCP integration.

There are  two break-out use-cases for AI in the commercial world:

- code completion and
- deep research.

Both excel in personal setups and enterprise environments alike.

And they’re the break-out use-cases for two straightforward reasons…

First off, they live in environments with massive, high-quality contextual reference data.

Give AI abundant, relevant contextual data, and it generates succinct and accurate responses.

Sure, you could prompt Grok to create a white paper on the upsides of XYZ tech — but without that depth contextual reference, you’ll likely end up with generic fluff, the kind of “work slop” that is not useful.

Deep Research Agents — autonomous LLM-based systems that emulate human analysts by iteratively retrieving evidence, decomposing problems and constructing higher-order insights.

Second, success hinges on human oversight.

Vast amounts of research back this…

Agentic AI thrives on a spectrum, where you dial autonomy up or down based on the task.

Full autonomy for straight forward work, tight human reins for more complex and critical tasks.

It’s not about all or nothing automation — it’s about smart collaboration.

I would also add that inference latency is a significant impediment for Conversational UIs based on LLMs.

For obvious reasons deep research is not subject to the demands of latency. Which helps the use-case as-well.

Considering that the finding of data, vetting and synthesising take a vast amount of time, this is actually a use-case where inference latency pales against human driven effort time scales.

According to CB Insights, The AI agent moment is reshaping enterprise software...since their last mapping in March 2025, the landscape has exploded from roughly 300 players to thousands as tech companies race to launch AI agent offerings across horizontal use cases and industry applications.

Agentic solutions have become a leading acquisition target for enterprise software incumbents, while 1 in 5 new unicorns ($1B+ valuation) are now developing agents.

To help navigate this expansion, CB Insights mapped 400+ promising private companies building AI agent applications.

𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀 𝗧𝗼 𝗕𝗮𝗹𝗮𝗻𝗰𝗲 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝘆 𝘄𝗶𝘁𝗵 𝗛𝘂𝗺𝗮𝗻 𝗢𝘃𝗲𝗿𝘀𝗶𝗴𝗵𝘁

Agentic workflows are changing how knowledge workers are interacting with technology.

Workflows use autonomous AI Agents to perform complex tasks with minimal human intervention, but oversight.

While AI Agents introduce a new level of autonomy, they still require strong human supervision to ensure accuracy and reliability.

What Are Agentic Workflows?

Agentic workflows refer to processes where AI Agents — powered by advanced large language models (LLMs) — execute multi-step tasks autonomously.

Unlike traditional LLMs that generate text in response to prompts, AI Agents can plan, reason and interact with external tools, such as APIs, databases, or web browsers, to achieve specific goals.

For example, an AI Agent might book a flight, draft an email, or scrape data from a website by breaking down the task into actionable steps.

Workflows leverage a more modular approach AI Agent frameworks, that integrate LLMs with toolkits for decision-making and task orchestration.

Two uses cases are really coming to the fore...

...and, data quality and integration are the biggest barriers to AI agent adoption...

The question is not adoption of AI agents but how to scale them strategically while addressing integration challenges (46%), data quality requirements (42%), and change management needs (39%)...

The 2026 State of AI Agents Report from Anthropic

A new open-source release from Salesforce AI Research: Enterprise Deep Research (EDR), a multi-agent system designed for automated, steerable deep data analysis in enterprise environments.

EDR breaks down complex research queries into manageable steps using a master planning agent, then leverages specialised search agents—for general web, academic papers, GitHub repos, and LinkedIn profiles—to gather targeted information. 

A reflection mechanism helps identify knowledge gaps and refine directions, with built-in support for human-in-the-loop steering.

Key highlights:
- Compatible with multiple LLM providers (OpenAI, Anthropic, Google, Groq, SambaNova).
- Includes a web app for real-time interaction, document analysis, and citation tracking.
- Backed by the EDR-200 dataset (200+ agentic research trajectories) on Hugging Face for benchmarking.
- Apache 2.0 licensed, with code, benchmarks, and a demo video available.

This framework aims to streamline data-driven decision-making in professional settings, from market analysis to competitive intelligence.

(Shelby Heinecke, PhD) 

Check out the repo here: https://lnkd.in/d-R4zvDs

And the accompanying paper: https://lnkd.in/dzp8XT26

Video source: https://lnkd.in/d-R4zvDs

From Handcrafted to AI-Driven Adaptive Workflows

For decades, businesses handcrafted rigid processes, mapping every step from start to finish.

Take enterprise chatbots: developers designed custom flows for every conceivable scenario. Each flow targeted a specific customer intent—once detected, a fixed, predetermined path kicked in.

This manual blueprinting of decision trees, dependencies, and execution sequences formed the core of early automation.

But it had fatal flaws: zero adaptability to shifting user behaviors or evolving products, plus endless maintenance to patch vulnerabilities and updates.

Enter AI: OpenAI's recent research shows reasoning models like o1 transforming static knowledge articles into dynamic, agentic workflows—sequences of events laced with smart conditions.

o1 excels here, deconstructing complex docs (multi-step guides, decision trees, even diagrams) into executable routines. Its zero-shot prowess means no heavy training or fine-tuning—just feed it the article, and it maps out clear, adaptive steps on the fly.

The result? AI isn't just automating tasks—it's architecting resilient, self-evolving workflows that scale with real-world chaos.

Find the detail here: https://lnkd.in/dbGXvs8f

Research found that the majority of AI Agent implementations are internal employee facing...a safe place to get stated. Just over 40% us external facing, a bit more tricky. 

And then Agent-to-Agent / Machine-to-machine AI agents represent just 7% of deployments — tempering the hype around expansive AI Agent meshes for the foreseeable future.

A recent study found the following reasons AI Agents are built...(see graph)

They also found that:

- 47% of AI Agents executing fewer than five steps.

- And 46.7% with less than 5 model calls before requiring human intervention. Tightly bounded autonomy.

- 70% rely on prompting off the-shelf models instead of weight tuning.

- 74% depend primarily on human evaluation, while 52% use LLM-as-a-judge.

- A staggering 85% of detailed case studies do not use third-party agent frameworks, opting instead to build custom agent application from scratch.

- 75% evaluate their agents without formal benchmarks, relying instead on online-tests such as A/B testing or direct expert/user feedback.

- Internal employees are the primary user base (52.2%) for agents. Which makes sense, the level of cooperation you will get from employees as apposed to customers. This is followed by external customers(40.3%).

- Only 7.5% of deployed systems serve non-human consumers.

There are agentic features released which are imbedded in new models, and leveraged via the SDK...

It feels to me that with OpenAI’s GPT-5.2, the concept of tighter coupling between model and runtime is emerging.

Powerful models are equipped with pre-defined tools (functions) which is defined via the SDK. 

(In this article I have a complete working example in Python you can copy and paste into Colab.)

This seems to be as a pivotal advancement in building more capable AI Agents by creating a thin framework of instructions around the model invocation.

xAI did something similar with their agentic server-side tool/function calling. 

Where tools are orchestrated in a sequence by the model in a true agentic fashion. 

For me agentic means that there is a certain level of freedom, agency, the system has.

As agents grow increasingly powerful…
- demonstrating superior long-horizon reasoning,
- reliable tool-calling, and
- end-to-end execution of complex tasks…

I guess the beauty of this approach is that the user defines their tools and the model, by leveraging the OpenAI SDK, plans accordingly.

GPT-5.2 addresses the limitations of previous loosely integrated systems by embedding domain-specific expertise more seamlessly into the core inference process.

This reduces latency, minimises errors in multi-step workflows, and enhances overall reliability.

For much of the past three years, the visible impact of AI has been most
apparent among consumers. 

More than 1 million business customers now use OpenAI’s tools.

01 Enterprise usage is scaling, wit h deeper workflow integration.

02 Enterprises t hat leverage AI are experiencing measurable productivity and business impact. 

03 Enterprise growth is global and rapidly accelerating across industries. 

04 A widening gap is emerging between leaders and laggards.

OpenAI: The state of enterprise AI

I found this guide from Perplexity on how to use AI at Work...I guess it's a veiled marketing exercise to some extent. But, it does give some insight on how AI can be used at work. 

Perplexity says:

"Work doesn’t have to feel overwhelming or scattered. By bringing together focused attention, smarter ways to scale your talents, and real steps toward measurable results, you turn busywork into progress and ideas into impact. 

When AI takes care of the friction, you’re free to do the work that moves you forward—thinking deeply, solving important problems, and creating something meaningful."

taplio

The Battle

William Szamosszegi

Cobus Greyling

Followers

Posts

Engagements

Likes

Performance ScoreWilliam Szamosszegi's LinkedIn Performance

Performance ScoreCobus Greyling's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

William Szamosszegi

Cobus Greyling

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);}William Szamosszegi's LinkedIn Performance

Performance ScoreCobus Greyling's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreWilliam Szamosszegi's LinkedIn Performance