Agentic Conversations (formally mlops.community)

The DuckLake Lakehouse Format // MLOps Podcast #339 with Hannes Mühleisen, Co-founder and CEO of DuckDB Labs.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

Managing data on Object Stores has been a painful affair. Users had to choose between data swamp chaos or a maze of metadata files with catalog servers on top.

DuckLake is a new paradigm for managing data on object stores: First, it uses classical SQL data management systems to manage metadata. Second, actual data is stored in Parquet files on pretty arbitrary storage. Third, processing queries is done client-side, or anywhere really. DuckDB is the first system to integrate with DuckLake using an extension with the same name.

Conceptually, DuckLake enables central control over truth while decentralizing compute and storage entirely. DuckLake turns data warehouse architecture upside down by departing from the integrated metadata/compute layer towards a fully disconnected operation with only centralized metadata. For the first time, DuckLake allows a “multi-player” experience with DuckDB, where computation stays fully local, but transactional control is centralized.

// Bio

Hannes Mühleisen 🔈 is a creator of the DuckDB database management system and Co-founder and CEO of DuckDB Labs. He is a senior researcher at the Centrum Wiskunde & Informatica (CWI) in Amsterdam. He is also Professor of Data Engineering at Radboud University Nijmegen.

// Related Links

Website: https://hannes.muehleisen.orgUnleashing Unconstrained News Knowledge Graphs to Combat Misinformation // Robert Caulk // #279 - https://youtu.be/pF8zTI867EI

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Hudson on LinkedIn: /hfmuehleisen

Timestamps:

[00:00] Spooky ease in tech

[00:29] DuckDB and DuckLake

[07:50] Pain vs trust factors

[13:12] Prioritizing project features

[16:16] Platform growth tension

[22:06] Building principles

[25:26] OSS vs system reliability

[30:27] Creative uses of DuckDB

[35:35] Tecton product strategy

[43:30] Mindset shift

[52:25] DuckDB future shifts

[55:37] Wrap up

AI Conversations Powered by Prosus Group 

Russ d'Sa shares how LiveKit went from a small open-source project during the pandemic to powering voice interfaces for giants like OpenAI. He talks about the turning point when LiveKit teamed up on ChatGPT’s voice features, the challenges of making AI sound human, and why voice could be the future of multimodal AI. It’s a story of chance, big shifts, and building the backbone of tomorrow’s AI.

Guest speakers: Russ Dsa - CEO of LiveKit

Host: Demetrios Brinkmann - Founder of MLOps Community

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

#voiceagents #podcast

AI Conversations Powered by Prosus Group. 

Craig Tavares, COO of Buzz High Performance Compute, shares lessons from building GPU cloud infrastructure worldwide. He stresses the role of sovereign mandates, renewable power, and modular cooling in scaling data centers, while highlighting NVIDIA partnerships and orchestration as key to sustaining AI workloads.

Guest speaker: Craig Tavares - COO of Buzz HPC

Host: Demetrios Brinkmann - Founder of MLOps Community

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Trust at Scale: Security and Governance for Open Source Models // MLOps Podcast #338 with Hudson Buzby, Solutions Architect at JFrog.

Appreciate JFrog for their support in bringing this blog to life.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

For better or for worse, machine learning has traditionally escaped the gaze of security and infrastructure teams, operating outside traditional DevOps practices and not always adhering to organizations' development or security standards. With the introduction of open source catalogs like HuggingFace and Ollama, a new standard has been established for locating, identifying, and deploying machine learning and AI models. But with this new standard comes a plethora of security, governance, and legal challenges that organizations need to address before they can comfortably allow developers to freely build and deploy ML/AI applications. In this conversation, we will discuss ways that enterprise-scale organizations are addressing these challenges to safely and securely build these development environments.

// Bio

Hudson Buzby is a solution engineer with an emphasis on MLOps, LLMOps, Big Data, and Distributed Systems, leveraging his expertise to help organizations optimize their machine learning operations and large language model deployments. His role involves providing technical solutions and guidance to enhance the efficiency and effectiveness of AI-driven projects.

// Related Links

https://www.youtube.com/channel/UCh2hNg76zo3d1qQqTWIQxDg

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Hudson on LinkedIn: /hudson-buzby/

Timestamps:

[00:00] Value of Centralized Gateway

[00:35] Point Break vs Big Lebowski

[01:47] AI adoption failure stats

[05:12] ML vs Generative AI

[12:04] LLM adoption in enterprise

[18:08] MLOps Community alternative

[23:43] AI governance challenges

[27:39] Organizational debt comparison

[31:41] AI tool sprawl

[35:59] MLOps to platform evolution

[40:56] MLOps then vs now

[49:48] Model trust and safety

[52:19] AI model effectiveness

[55:54] Product discovery process

[58:38] Wrap up

AI Conversations Powered by Prosus Group 

Nishikant Dhanuka talks about what it really takes to make AI agents useful—especially in e-commerce and productivity. From making them smarter with context (like user history and real-time data) to mixing chat and UI for smoother interactions, he breaks down what’s working and what’s not. He also shares why evals matter, how to test with real users, and why AI only succeeds when it actually makes life easier, not more complicated.

Guest speaker: Nishikant Dhanuka - Senior Director of AI at Prosus Group

Host: Demetrios Brinkmann - Founder of MLOps Community

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

#contextengineering #aiengineer #aiinfrastructure #podcast

The Era of AI Agents in Marketing // MLOps Podcast #337 with Joel Horwitz, Growth Engineer at Neoteric3D.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

We’re entering a new era in marketing—one powered by AI agents, not just analysts. The rise of tools like Clay, Karrot.ai, 6sense, and Mutiny is reshaping how go-to-market (GTM) teams operate, making room for a new kind of operator: the GTM engineer. This hybrid role blends technical fluency with growth strategy, leveraging APIs, automation, and AI to orchestrate hyper-personalized, scalable campaigns. No longer just marketers, today’s GTM teams are builders—connecting data, deploying agents, and fine-tuning workflows in real time to meet buyers where they are. This shift isn’t just evolution—it’s a replatforming of the entire GTM function.

// Bio

Joel S. Horwitz has been riding the data wave since before it was cool—literally. He spoke at Spark Summit back in 2014 and penned a prescient piece for MIT Tech Review on data science and machine learning before they became boardroom buzzwords. A former big tech executive turned entrepreneur, Joel now runs Neoteric3D (N3D for short), a digital design and data growth agency that helps brands scale with smarts and style. When he’s not architecting next-gen growth strategies, you’ll find him logging long miles on the trail or coaching his sons’ soccer and baseball teams like a champ.

// Related Links

Website: https://www.neoteric3d.com

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Joel on LinkedIn: /joelshorwitz

Timestamps:

[00:00] Joel's preferred coffee

[00:53] Agentic workflows in marketing

[04:26] Agentic AI vs big data

[08:24] Creative outreach automation

[13:08] LLMs in marketing optimization

[17:36] Traffic relevance

[23:36] End-to-end AI workflow

[28:10] AI in task automation

[32:08] AI systems architecting

[38:00] AI vs Thought Leadership

[43:10] AI as sparring partner

[45:22] AI shifts human roles

[48:23] Wrap up

Distilling 200+ Hours of NeurIPS: What’s Next for AI // MLOps Podcast #336 with Nikolaos Vasiloglou, VP of Research ML at RelationalAI.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

Nikolaos' widely shared analysis on LinkedIn highlighted key insights across agentic AI, scaling laws, LLM development, and more. Now, he’s exploring how AI itself might be trained to automate this process in the future, offering a glimpse into how researchers could harness LLMs to synthesize conferences like NeurIPS in real-time.

// Bio

Nikolaos Vasiloglou is VP of Research-ML for RelationalAI, the industry's first knowledge graph coprocessor for the data cloud. Nikolaos has over 20 years of experience implementing high-value machine learning and AI solutions across various industries.

// Related Links

Website: https://relational.ai/

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Nikolaos on LinkedIn: /vasiloglou/

Timestamps:

[00:00] Nik's preferred coffee

[01:05] Distilling NeurIPS insights

[06:43] Choosing research papers

[16:49] Agent patterns at NeurIPS

[21:16] Interest in agent-based innovation

[25:54] Time series forecasting models

[28:15] Tabular foundation models

[36:25] Verifier challenges and complexity

[39:36] Knowledge graph

[45:00] Knowledge graph data challenges

[47:14] Worldview in knowledge graphs

[50:30] Self-serve analytics challenges

[56:22] Llama model adaptation comparison

[56:59] Wrap up

AI Conversations Powered by Prosus Group 

Demetrios chats with Beyang Liu about Sourcegraph’s AMP, exploring how AI coding agents are reshaping development—from IDEs to natural language commands—boosting productivity, cutting costs, and redefining how developers work with code.

Guest speaker:

Beyang Liu - CTO of Sourcegraph

Host:

Demetrios Brinkmann - Founder of MLOps Community

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

AI Conversations Powered by Prosus Group 

Stack Overflow is adapting to the AI era by licensing its trusted Q&A corpus, expanding into discussions and enterprise tools, and reinforcing its role as a reliable source as developer trust in AI output declines.

Guest speaker:

Prashanth Chandrashekar - CEO of Stack Overflow

Host:

Demetrios Brinkmann - Founder of MLOps Community

2025 Developer Survey: https://survey.stackoverflow.co/2025?utm_medium=referral&utm_source=direct-share&utm_campaign=dev-survey-2025&utm_content=MLOps

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Knowledge is Eventually Consistent // MLOps Podcast #335 with Devin Stein, CEO of Dosu.

Grateful to  @Databricks  and  @hyperbolic-labs  for supporting our podcast and helping us keep great conversations going.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

AI as a partner in building richer, more accessible written knowledge—so communities and teams can thrive, endure, and expand their reach.

// Bio

Devin is the CEO and Founder of Dosu. Prior to Dosu, Devin was an early engineer and leader at various startups. Outside of work, he is an active open source contributor and maintainer.

// Related Links

Website: https://github.com/devstein

https://www.youtube.com/watch?v=sC8aW47DqPg

https://www.youtube.com/watch?v=PuM0Gd3txfQ

https://www.youtube.com/watch?v=ah6diDQ9wyw

https://www.youtube.com/watch?v=x22FEQic8lg

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our Slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Devin on LinkedIn: /devstein/

Timestamps:

[00:00] Devin's preferred coffee

[00:53] Facts agent overview

[03:47] Decision state detection

[07:55 - 8:41] Databricks ad

[08:42] Context-dependent word meanings

[15:25] Fact lifecycle management

[24:40] Maintaining quality documentation

[30:10 - 31:06] Hyperbolic ad

[31:07] Agent collaboration scenarios

[38:22] Knowledge maintenance

[44:10] Deployment and integration strategies

[48:13] Flywheel data approach

[51:54] Horror story engineering function

[54:32] Wrap up