Aurimas Griciūnas

Senior Solutions Architect

SwirlAI

Driving MLOps, Data Engineering and Machine Learning industry forward by helping professionals Level Up in these disciplines.

Founder & CEO

neptune.ai

Chief Product Officer

Head of Product

Driving MLOps industry forward by helping Neptune customers succeed in adopting MLOps best practices.

Transactie Monitoring Nederland (TMNL)

- Contributed to the development of end-to-end Machine Learning platform capabilities in a highly secure environment.
 - Focused on leveraging AWS SageMaker and supporting AWS services to enable scalable experimentation and model deployment.
 - Collaborated with the team to build reusable template libraries, streamlining the process of running ML experiments and deploying models efficiently and securely.

Senior MLOps Engineer

HomeToGo

- Led cross-functional teams of Cloud Engineers and Data Platform Engineers, driving key infrastructure and data initiatives.
 - Spearheaded a company-wide Cloud-Native Transformation initiative to modernize applications and infrastructure.
 - Championed the migration to Infrastructure as Code (IaC), standardizing infrastructure provisioning using Terraform for improved scalability, consistency, and automation.

Data Architect / Data Platform & Cloud Engineering Lead

- Led a Team of Data Engineers.
- Led a Team of Data Platform Engineers.
- Initiated and led a successful company wide Cloud Native Transformation.
- Migrated entire Data Platform Stack to Kubernetes. This included:

        - Airflow.
        - Superset.
        - Spark.
        - MLFlow.
        - Flink.
        - Airbyte.
        - DataHub.

- Introduced first Stream Processing applications that moved some of the Data Warehouse Processing stages upstream the pipeline.

Senior Data Engineer / Data Platform Team Lead

- Led a team of Data Engineers.
- Initiated and led a Data Warehouse migration from Redshift to Snowflake effort. It also included a refactoring of the Data Model.
- Oversaw Data migration from a Druid cluster to Redshift and later Snowflake.
- Introduced and rolled out Machine Learning Experiment Tracking System.
- Deployed and maintained first on-prem Kubernetes Clusters in the company.

Senior Data Engineer / Team Lead

- Led a successful implementation and optimisation of a custom Website Activity Tracking Ingestion Pipeline.
- Deployed and rolled out a company-wide Airflow Cluster.
- Led a full migration from Jenkins to Airflow for Data Engineering Pipelines.

Senior Data Engineer

Tesonet

Responsible for end-to-end development of ML-based products and big data solutions:

 - Identified business use cases suitable for machine learning and translated them into technical solutions.
 - Designed and implemented the full ML lifecycle, including data extraction, cleaning, validation, model training, packaging, and deployment.
 - Contributed to the architecture and development of scalable big data ingestion pipelines.
 - Set up and configured Kafka topics; developed real-time processing applications using Kafka Streams and Spark.

Senior Data Scientist/Engineer

Danske Bank

Projects I worked on:

-Affordability model enhancement using transaction data.
-Bringing transaction data into Balance Sheet stress-testing.
-Unlocking potential of transaction data throughout Risk Management function.
-Stress-testing of REA components and dynamic analysis on different credit portfolio levels.
-Process optimization and automation. Preparing models for production placement.

Senior Data Scientist/Analyst

Projects I worked on:

-Retail customer affordability project in Big Data analytics team.
-IFRS9 Life-time probability of default (LTPD) model development.
-IFRS9 Prepayment model development.
-IFRS9 impact on P&L analysis.
-IFRS9 and pre-IFRS9 ECL stress-testing and reporting.
-REA stress-testing and dynamic analysis on different credit portfolio levels.
-Process optimization by both improving old approaches and introducing new technologies.

Analyst

ADB Gjensidige

Was responsible for implementation of Solvency2 regulatory standards and providing required reports.

Actuary

Was responsible for generating business analysis reports for management.

Was responsible for calculation of Technical Reserves for Non-Life insurance products by applying different triangulation methods.

Junior Actuary

Follow me to learn about Data Engineering and ML Systems | Author of SwirlAI Newsletter | Public Speaker

Tired of all the noise surrounding AI space in recent years? Having a hard time keeping up with all of the news in the field and distinguishing what is really important? Follow me and subscribe to my newsletter to get actionable insights on building production ready AI Systems and keep track of only the important signal of what is happening in the AI world.

With over a decade of experience in Data Science, AI/ML Engineering, and MLOps, I’m also dedicated to helping practitioners level up their skills in MLOps, Data Engineering, Machine Learning, and related fields. I share my knowledge in a clear and accessible way to make learning straightforward and actionable.

If you want to partner with SwirlAI and get your product in front of 40K+ AI/ML/Data Engineers and Leaders via our Newsletter or 135K+ AI/ML/Data Engineers and Leaders via LinkedIn, contact me directly.

Leading Product Strategy.

TOON (Token-Oriented Object Notation) is out for some days now and it aims to make communication with LLMs more accurate and token-efficient.

The TOON topic is now one of the hottest news on the LLM market and it might actually matter.

𝗪𝗵𝘆 𝗜 𝘁𝗵𝗶𝗻𝗸 𝘀𝗼:

I was initially hesitant to cover this, potentially being another hype to quickly fade, but:

✅ The format has been shown to increase the accuracy of models while decreasing the token count. I was not sure if there were any accuracy retention studies made, it seems there were.
✅ Token efficiency is extremely important when working with Agentic Systems that require a lot of structured context inside of their reasoning chains. And we are moving towards a post-PoC world where there is a lot of emphasis placed on optimisation of the workflows.

𝗔 𝘀𝗵𝗼𝗿𝘁 𝘀𝘂𝗺𝗺𝗮𝗿𝘆:

- Token-efficient: typically 30-60% fewer tokens on large uniform arrays vs formatted JSON.
- LLM-friendly guardrails: explicit lengths and fields enable validation.
- Minimal syntax: removes redundant punctuation (braces, brackets, most quotes).
- Indentation-based structure: like YAML, uses whitespace instead of braces.
- Tabular arrays: declare keys once, stream data as rows.

An example:

𝘑𝘚𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵:

"shopping_cart": [
    { "id": "GDKVEG984", "name": "iPhone 15 Pro Max", "quantity": 2, "price": 1499.99, "category": "Electronics" },
    { "id": "GDKVEG985", "name": "Samsung Galaxy S24 Ultra", "quantity": 1, "price": 1299.99, "category": "Electronics" },
    { "id": "GDKVEG986", "name": "Apple Watch Series 9", "quantity": 1, "price": 199.99, "category": "Electronics" },
    { "id": "GDKVEG987", "name": "MacBook Pro 16-inch", "quantity": 1, "price": 2499.99, "category": "Electronics" }
  ]
}

𝘞𝘩𝘦𝘯 𝘦𝘯𝘤𝘰𝘥𝘦𝘥 𝘪𝘯𝘵𝘰 𝘛𝘖𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵:

shopping_cart: items[4]{id,name,quantity,price,category}:
    GDKVEG984,iPhone 15 Pro Max,2,1499.99,Electronics
    GDKVEG985,Samsung Galaxy S24 Ultra,1,1299.99,Electronics
    GDKVEG986,Apple Watch Series 9,1,199.99,Electronics
    GDKVEG987,MacBook Pro 16-inch,1,2499.99,Electronics

𝗥𝗲𝘀𝘂𝗹𝘁:

✅ 43% savings in token amount.
✅ Directly translates to 43% savings in token cost for this LLM input.

❗️ Be sure to know when NOT to use the format (and always test it for your application specifically):

- Deeply nested or non-uniform structures.
- Semi-uniform arrays.
- Pure tabular data.

ℹ️ I will be testing it in the upcoming weeks.

Let me know if you have already tested TOON and what are your takeaways! 👇

Founder @ SwirlAI • UpSkilling the Next Generation of AI Talent • Author of SwirlAI Newsletter • Public Speaker

𝗠𝗖𝗣 plus 𝗔𝟮𝗔, here is how they complement each other 👇

Protocol wars continue to rage, let's understand how Googles A2A (Agent2Agent) protocol is different from MCP and how they complement each other (read till the end).

𝘔𝘰𝘷𝘪𝘯𝘨 𝘱𝘪𝘦𝘤𝘦𝘴 𝘪𝘯 𝘔𝘊𝘗:

𝟭. MCP Host - Programs using LLMs at the core that want to access data through MCP.

❗️ When combined with A2A, an Agent becomes MCP Host.

𝟮. MCP Client - Clients that maintain 1:1 connections with servers.
𝟯. MCP Server - Lightweight programs that each expose specific capabilities through the standardised Model Context Protocol.
𝟰. Local Data Sources - Your computer’s files, databases, and services that MCP servers can securely access.
𝟱. Remote Data Sources - External systems available over the internet (e.g., through APIs) that MCP servers can connect to.

𝘌𝘯𝘵𝘦𝘳 𝘈2𝘈:

Where MCP falls short, A2A tries to help. In multi-Agent applications where state is not necessarily shared

𝟲. Agents (MCP Hosts) would implement and communicate via A2A protocol, that enables:

➡️ Secure Collaboration.
➡️ Task and State Management.
➡️ User Experience Negotiation.
➡️ Capability discovery - similar to MCP tools.

𝗛𝗼𝗻𝗲𝘀𝘁 𝘁𝗵𝗼𝘂𝗴𝗵𝘁𝘀:

❗️ Open protocols for Agent communication are important but they are just a piece of the picture, there will be a need for standards that govern all of the existing protocols and other missing pieces. E.g.

❓ How do we standardise tracing and Observability in multi-agent IOA systems?
❓ How do we retain the identity of the running job of an AI Agent instance if the communication standard is not unified in different parts of the pipeline?
❓ ... 

✅ More on this in my future posts, stay tuned!

Let me know your thoughts in the comments. 👇

Snowflake

Principal Solutions Architect

Qlik

Principal PreSales Solutions Architect - Analytics / Business Intelligence

I work closely with sales teams (both local & international) to provide product demonstrations, web and windows based art of the possible proof of concepts and most importantly understanding and solving real world business problems using data and analytics. Major part of this role is to be able to understand business problems that the clients are experiencing and to solve their problems with creative & innovative solutions by using business intelligence/visualizations & data analysis techniques. Final challenge is to effectively demonstrate these solutions to both technical & executive teams to have their buy in. 

Best part of this role is to have the opportunity to work with large enterprise customers from wide variety of business verticals and to be able to solve their problems using data by showing them insights that they have never seen before and seeing the excitement that generates  .

Senior PreSales Solutions Architect - Business Intelligence

Helping commercial and large enterprise level organizations achieve strategic insights about their businesses by using the latest technologies in business intelligence and visual analytics. Proving presales support by software demonstrations, technical deep dives and designing BI proof of concept solutions which allow businesses to analyze and discover trends and hidden insights within their data by blending data from disparate sources.

Enterprise PreSales Solutions Architect

Metafile Information Systems, Inc.

Director of Presales Engineering

As a solutions architect specialized in document management, imaging & workflow, I was responsible for sales support as well as designing and implementing ECM/workflow solutions to mid to large size companies integrating automated document management systems with various ERP solutions such as Dynamics (GP, AX, Nav, SL), Oracle, SAP, Sage, Lawson & etc. 

As part of this role I worked closely with our internal sales team and customers who were primarily managers & executives in HR, AP, AR, Marketing, Sales, Finance departments to understand their business processes and  develop an automated Workflow/Imaging/Content Management system that will meet their goals and increase the productivity.
support.

I also created numerous private and public demo assets that were used within the presales teams to increase sales.

Sr. Pre-Sales Solutions Architect 

Supported number of sales teams with technical demonstrations, RFP/RFI responses & technical proof of concepts for creating workflow/approval automation systems in wide variety of departmental solutions such as Sales, Marketing, HR & etc.

Supprted and was the lead solutions architect over 120 new new accounts.

Pre-sales Solutions Architect

As a developer I created ready to deploy AP document management/workflow solutions integrated directly with Microsoft GP, SL ERP systems as well as NetSuite.

These preconfigured solutions saved both time and money in terms of customer solution deployments by eliminating a lot of technical and manual processes as well as replacing API based integration techniques to automatically create transactions in these ERP systems.

Skills utilized for this role: 

Integration Technologies: T-SQL procedures, SOAP/REST web services, Business Object APIs, XML & Flat file transfers. 

Technical writing: Creating technical documentation including project implementation plans, system/workflow diagrams, statements of work, PowerPoint presentations.

Database: Extensive experience with Ms SQL Server database design and administration. Also proficient with Sybase as well.
 
Programming: VB.NET / ASP.Net / VB6, Java & Javascript. Creating custom applications ranging from data conversion tools to 3rd party integration projects.

Developer/Systems Integrator

I was responsible for implementing, configuring solutions on customer sites. This required extensive technical skills including database systems, SQL, VB programming and various Ms Windows administrative functions.

As part of this role, I was also tasked with providing custom training classes to customers as well as creating support & help documentation specific to their implemented solutions.

Technical Lead

Cloud, Data, Analytics & BI expert experienced in helping organizations gain positive business outcomes through the use of modern analytics. Pre-sales architect with over 18 years of experience assisting software sales in large enterprise sales cycles by creating innovative demos, developing prototype applications and creating solutions to address complex data related challenges in the field of business intelligence(BI) & Big data analytic & business process automation, imaging and workflow.

Nick Akincilar

For anyone questioning Snowflake’s #ML power — after Coinbase, PayPal is now showcasing how our #distributed #Snowpark ML training with #XGBoost, #FeatureStore, #ModelRegistry and batch inferencing can scale to predict merchant fees across hundreds of millions of credit card transactions.

If you rewind back 33mins, you can see #AI_SQL for product damage image detection, damage summary as well as online feature store and real time online store recommendation ML models in action.

WATCH PAYPAL BUILD KEYNOTE SEGMENT HERE
https://lnkd.in/eSxk2Mga

Principal Architect | GenAI | Data Strategist | Cloud | Tech Writer

Snowflake Horizon #Iceberg Rest Catalog(IRC) is now in Public Preview. You now have a single catalog (one set of users, roles, authentication) that governs both standard & iceberg tables.  

This allows any 3rd party Query Engine capable of connecting to catalogs using IRC Rest endpoints to authenticate & query Snowflake managed Iceberg tables directly.

Docs:
https://lnkd.in/eq7ipKxh

Blog & Tutorials 
https://lnkd.in/eud8jgwq


Why Snowflake supporting IRC or IRC in general is important?

SECURITY:
- External engines only need a REST endpoint & authentication token. #Snowflake IRC supports Enterprise authentication options such as External Oauth, Key-Pair Authentication, Programmatic Access Token (PAT).  Provider catalog does the authorization & RBAC. 

- Credential vending: This means engine never needs to access the root of the underlying object store where the files reside. Using short-lived credentials reduces the risk of leaks and are intentionally down-scoped to the minimum permissions necessary (e.g., read/write only) for the exact files (not the root or folders) that the query needs to access.  This avoids being forced to grant permanent full access to your iceberg table storage bucket root folder that has data for ALL tables just to be able to share ONE table.

- Private Link support for #AWS and #Azure (#GCP coming soon)

GOVERNANCE & MANAGEMENT:
- One catalog for both standard & iceberg tables, One set of users, One set of roles & RBAC means consistent governance. Allows you to extend Snowflake Horizon RBAC (identities/roles) beyond your Snowflake environment with other engines.

BETTER COST & PERFORMANCE:
Since the engines & catalogs communicate through standard #OSS REST protocol, catalogs automatically deliver required #metadata as the engine traverses the catalog.

This means a 3rd party engine DOES NOT NEED:
1. Additional connectivity layer using JDBC to the provider platform
2. to run slew of SQL queries each time(SELECT databases, SELECT tables, SELECT columns) prior to running any query against iceberg tables using this JDBC connection (slow & inefficient)
3. Use providers compute to execute these catalog queries (extra cost for provider)

EASE OF USE:
Since the only thing needed is the catalog rest endpoint + an access token, it makes things extremely simple both for the provider as well as the 3rd party engine.
 
As most other popular query engines & platforms such as #Fabric, #BigQuery, #Spark, #Trino, #EMR, #DuckDB & etc. have built-in support to leverage IRC catalogs to query iceberg tables, this allows these engines to securely & easily leverage data managed by Snowflake. 

One notable exception is #databricks. It shares its internal tables via IRC(makes it easy for others to read if data is in Unity), but won't allow their users to connect to other iceberg catalogs via IRC by forcing a complex, difficult & insecure workaround. (Makes it difficult to read if data is Not in Unity)

A very common question I get is, 'Why choose Cortex AI Analyst & Snowflake Intelligence over other solutions?' The simple answer: #ACCURACY & #SPEED! 

When leveraging #AI with your data, accuracy is king. Snowflake invests thousands of engineering hours to ensure our solutions are the most accurate and fastest on the market, so you don't have to - and can simply skip the struggle.

This dedication translates to faster time-to-value for our customers. They're building production-ready, AI-driven #data applications with the necessary accuracy, #security, #governance, and #scalability to empower both the #business and #consumers.

Check out this deep-dive from our engineering team that explains just one of the key ways we make Cortex Analyst Service a superior, production-ready Text-2-SQL solution. (Thanks, Jonathan Regenstein, for sharing this gem!)

https://lnkd.in/ep5fQFA6

Snowflake crushing #price, #performance & #TCO. Pricing is just out for the new #Interactive Warehouses and it is 40% cheaper than Gen1 & 55% cheaper than Gen2 compute at only 0.6 credits per hour while delivering 3X to 4X faster performance & throughput required for embedded in-product #analytics.

These multi-tenant in-product workloads require #high_concurrency & #low_latency queries that are typically highly selective, filtering #data based on such columns such as as #tenant_id or #customer_id.

Link to our pricing sheet with the new pricing for interactive analytcis. 
https://lnkd.in/ezr9CPMP

Just tested the new Snowflake #Interactive #Analytics feature, and the results are genuinely blowing the roof off the industry standard for #high_concurrency & #low_latency workloads!

This is a massive leap forward for building embedded analytics in customer-facing portals and products.

I immediately put it to the test with a heavy workload using a single XS instance of each with no multi-cluster option: 
- 1 billion-row Fact table with two dimensional joins (2,500 rows stores and 10,000 rows products).

The performance uplift is massive:

- 2.6X faster than the already incredibly quick Gen2 XS warehouse.
- 3.5X faster than the Gen1 XS warehouse.

THESE NUMBERS ARE JUST FOR 1 XS. If I had multiples running in parallel, it would be ton more but my laptop can't handle the volume.

Why This Matters:
Gen1 and Gen2 warehouses were already leading the pack for speed and concurrency, often only rivaled by expensive in-memory solutions. 

But with these new speeds, the need for costly, third-party low-latency point solutions for in-product and customer-facing analytics is effectively gone.

This new combo can easily run thousands of sub-second queries, finally making Snowflake the single, scalable source of truth for all your enterprise and embedded analytics needs.

Will share the dataset, queries & the Apache Jmeter test template soon.

#Snowflake #InteractiveAnalytics #DataWarehousing #EmbeddedAnalytics #LowLatency

Link to my original #Interactive #Analytics product announcement
https://lnkd.in/eVDTcHuw

taplio

The Battle

Aurimas Griciūnas

Nick Akincilar

Followers

Posts

Engagements

Likes

Performance ScoreAurimas Griciūnas's LinkedIn Performance

Performance ScoreNick Akincilar's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Aurimas Griciūnas

Nick Akincilar

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);}Aurimas Griciūnas's LinkedIn Performance

Performance ScoreNick Akincilar's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAurimas Griciūnas's LinkedIn Performance