Agentic Conversations (formally mlops.community)

Latest Episodes:

Building Effective Products with GenAI // Faizaan Charania // #187 Nov 03, 2023

Show notes

MLOps podcast #187 with Faizaan Charania, Product Manager, AI at LinkedIn, Building Effective Products with GenAI.

// Abstract

Faizaan outlines his AI product development approach, starting broadly and refining details with tech leads, emphasizing the value of a simplified MVP. He also explores integrating generative AI, highlighting its role in enhancing user experiences through LLMs.In this discussion, Faizaan shares wisdom on feedback integration, user trust, and the collaboration challenges between product managers and AI teams. Let's delve into evaluating AI-driven experiences and the complexities that arise in this dynamic landscape!

// Bio

Faizaan is an AI Product Lead at LinkedIn, working on Personalization and Generative AI use cases for Creators and Conversations on LinkedIn. He's been in the field of machine learning for 8+ years now. He started as a research assistant, eventually transitioning to being a Product Manager during his time at Yahoo.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

⁠--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Faizaan on LinkedIn: https://www.linkedin.com/in/faizaan-charania/

Timestamps:

[00:00] Faizaan's preferred beverage

[00:22] Takeaways

[02:54] The Bollywood actor

[05:23] Faizaan's in tech

[07:45] Technical pieces to learn about before working at LinkedIn

[09:23] Tech Team Data Strategy

[12:01] Gradual vs. Advanced ML Implementation

[13:36] Shipping on time

[14:11] Thoughts on building products with AI

[18:20] Push and pull mechanism

[21:47] Costs and Choices with AI Models

[25:06] AI ROI Evaluation

[27:02] Thoughts on open source

[28:17] Building Generative AI focus

[31:50] Prompts and Anomalies

[34:57] Where to have a human in the loop

[35:45] Problem-driven AI Tool

[37:56] Creator of AI-generated post on LinkedIn

[39:50] Product Impact on AI Democratization

[41:15] Distinct signals to measure ROI

[44:38] PMs learning AI while ML teams learn product

[47:22] Gotchas seen when adding a new AI feature

[50:00] Evaluation Challenges in Responses

[51:55] Who's more confident?

[52:55] Wrap up

The Future of Feature Stores and Platforms // Mike Del Balso & Josh Wills // # 186 Oct 31, 2023

Show notes

MLOps podcast #186 with Mike Del Balso, CEO & Co-founder of Tecton and Josh Wills, Angel Investor, The Future of Feature Stores and Platforms.

// Abstract

Mike and Josh discuss creating templates and working at a detailed level, exploring Tecton's potential for sharing fraud and third-party features. They focus on technical aspects like data handling and optimizing models, emphasizing the significance of quality data for AI systems and the necessity for cohesive feature infrastructure in reaching production stages.

// Bio

Mike Del Balso

Mike is the co-founder of Tecton, where he is focused on building next-generation data infrastructure for Operational ML. Before Tecton, Mike was the PM lead for the Uber Michelangelo ML platform. He was also a product manager at Google, where he managed the core ML systems that power Google’s Search Ads business.

Josh Wills

Josh Wills is an angel investor specializing in data and machine learning infrastructure. He was formerly the head of data engineering at Slack, the director of data science at Cloudera, and a software engineer at Google.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links⁠

--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Mike on LinkedIn: https://www.linkedin.com/in/michaeldelbalso/

Connect with Josh on LinkedIn: https://www.linkedin.com/in/josh-wills-13882b/

Timestamps:

[00:00] Introduction to Mike

[01:45] Takeaways

[03:32] Features of the new paradigm of ML and LLMs

[06:00] D. Sculley's papers

[13:05] The birth of Feature Store

[17:06] Data Pipeline Challenges Addressed

[20:00] Operationalizing

[26:50] Feature Store Challenges

[30:26] Z access

[36:23] Addressing Technical Debt Challenges

[37:27] Real-Time vs. Batch Processing

[47:10] Feature Store Evolution: Apache Iceberg

[49:59] Feature Platform: Dedicated Query Engine

[54:04] The bottleneck

[56:00] LLMs, Feature Stores Overview

[1:00:20] Vector databases

[1:06:15] Workflow Templating Efficiency

[1:08:35] Gamification suggestion for Tecton

[1:10:25] Wrap up

Lessons on Data Science Leadership // Luigi Patruno // #185 Oct 27, 2023

Show notes

MLOps podcast #185 with Luigi Patruno, VP of Data Science at 2U, Inc., Lessons on Data Science Leadership.

// Abstract

Picture this: you've got data products to manage, and you're in charge of a team. It's not all sunshine and rainbows, right? Luigi dives into the nitty-gritty of the challenges - from juggling data projects to wrangling the team dynamics. It's a real adventure, let me tell you!

// Bio

Luigi Patruno is a results-driven data science leader passionate about identifying value-add business opportunities and converting these into analytical solutions that deliver measurable business outcomes. As a leader, he focuses on defining strategic vision and, through motivation and discipline, driving teams of highly quantitative data scientists, machine learning engineers, and product managers to achieve extraordinary results. He is currently the VP of Data Science at 2U, where he leads the data science department focused on optimizing business operations through advanced analytics, experimentation, and machine learning. He enjoys teaching others how to leverage data science to improve their businesses through public speaking, teaching courses, and writing online at MLinProduction.com.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

Website: https://mlinproduction.com/

YouTube channel: https://www.youtube.com/playlist?list=PLBLnN4jzkyqkjLIRpDNZcsG7TMMEk9Asa

High Output Management book by Andrew Grove: https://www.amazon.nl/-/en/Andrew-S-Grove/dp/0679762884

The One Minute Manager by Kenneth Blanchard, Ph.D., and Spencer Johnson, M.D.: https://www.amazon.com/Minute-Manager-Kenneth-Blanchard-Ph-D/dp/074350917X

⁠--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Luigi on LinkedIn: https://www.linkedin.com/in/luigipatruno91/

Timestamps:

[00:00] Luigi's preferred coffee

[00:30] Takeaways

[03:04] Being practical

[05:44] Data-Driven Decision-Making in Management

[12:53] Recent Team Win

[14:43] The perfect storm

[20:22] Change Management and ROI

[25:09] Change Management: Navigating Resistance

[29:59] Clarifying North Star Communication

[36:24] OKRs in Data Science

[40:47] Success Likelihood in Business

[45:08] Bus problem solution

[49:25] Data Science-Platform Collaboration

[53:19] Decentralized Platforms Explained

[54:38] Data Platform Architecture Overview

[57:14] Incentives for Team Motivation

[1:09:45] The blind spots

[1:12:22] Wrap up

Data Platforms in MLOps: Translating Business Goals into Product Decisions // Richa Sachdev // #184 Oct 24, 2023

Show notes

MLOps podcast #184 with Richa Sachdev, Executive Director- Data Operations and Automation at JP Morgan Chase, Data Platforms in MLOps: Translating Business Goals into Product Decisions.

// Abstract

Richa, with her background in software engineering and experience in the financial sector, shares her insights on optimizing the end-user experience and the importance of understanding business goals and metrics. She discusses her journey in converting legacy applications, working with data platforms, and the challenges of integrating different databases. Richa also explores the role of automation in streamlining processes and improving customer interactions in the reward space. Join us as we unravel the fascinating world of MLOps and uncover the strategies and technologies that drive success in this ever-evolving field.

// Bio

A passionate and impact-driven leader whose expertise spans leading teams, architecting ML and data-intensive applications, and driving enterprise data strategy. Richa has worked for a Tier A Start-up developing feature platforms and in financial companies, leading ML Engineering teams to drive data-driven business decisions. Richa enjoys reading technical blogs focused on system design and plays an active role in the MLOps Community.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

https://www.youtube.com/watch?v=i0To3DeHGuU
https://www.youtube.com/watch?v=tAOf2lVQUY4
https://www.youtube.com/watch?v=cXanVyaannQ
https://www.youtube.com/watch?v=2aWSsL24fv8

--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Richa on LinkedIn: https://www.linkedin.com/in/richasachdev/

Timestamps:

[00:00] Richa's preferred coffee

[02:09] Takeaways

[04:26] Richa's background in data

[08:55] Prescriptive, Descriptive, and Predictive Data

[11:50] Data Engineering Perspectives & Setup

[17:34] Structured and Unstructured data

[21:01] Richa's day-to-day at Chase

[23:52] Figure out the business needs before the cool tech

[26:46] Importance of business metrics

[30:43] Optimizing end-user experience and trade-offs

[36:06] Exhausting creativity in finding solutions

[37:40] Consider faster implementation and increased ROI

[40:20] Banks still using COBOL

[41:17] Learning and growing as a versatile leader

[42:04] Wrap up

MLOps vs ML Orchestration // Ketan Umare // #183 Oct 20, 2023

Show notes

MLOps podcast #183 with Ketan Umare, CEO of Union.AI, MLOps vs ML Orchestration, co-hosted by Stephen Batifol.

// Abstract

Let's explore the relationship between Union and Flyte, emphasizing the significance of community-driven development and the challenge of balancing feature requests with security considerations. This conversation highlights the importance of real-time data and secure data handling in orchestrating machine learning models. The Flyte community's empathy and support for newcomers underscore the community's value in democratizing machine learning, making it more accessible and efficient for a broader audience.

// Bio

Ketan Umare is the CEO and co-founder at Union.ai. Previously, he had multiple Senior roles at Lyft, Oracle, and Amazon ranging from Cloud, distributed storage, Mapping (map-making), and machine-learning systems. He is passionate about building software that makes engineers' lives easier and provides simplified access to large-scale systems. Besides software, he is a proud father and husband, and enjoys traveling and outdoor activities.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

Website: https://union.ai/

Flyte: https://flyte.org/ ⁠

--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn:

https://www.linkedin.com/in/dpbrinkm/

Connect with Stephen on LinkedIn: https://www.linkedin.com/in/stephen-batifol/

Connect with Ketan on LinkedIn: https://www.linkedin.com/in/ketanumare/

Timestamps:

[00:00] Ketan's preferred coffee

[01:05] Takeaways

[03:08] Please like, share, and subscribe to our MLOps channels!

[03:15] Shout out to Ketan and UnionAI for sponsoring this episode!

[04:23] Orchestration of recent changes

[07:51] Community with Flyte

[11:26] ML orchestration

[15:40] 50/50 is generous

[20:06] Real-time ML

[21:15] Over-engineering without benefits

[23:20] Balancing everything

[27:40] Union verse Flyte

[32:52] High-value features of Union AI at the back of Flyte

[40:18] Building LLM infrastructure

[45:30] Traditional ML is the whole prompting

[46:46] LLMs for evaluating prompts

[48:55] Wrap up

MLOps@GetYourGuide // Jean Machado, Meghana Satish, Olivia Houghton, Theodore Meynard// #182 Oct 20, 2023

Show notes

MLOps podcast #182 with GetYourGuide's Jean Machado, DataScience Manager, Meghana Satish, MLOps Engineer, Olivia Houghton, Machine Learning Operations Engineer, Theodore Meynard, Data Science Manager, MLOps@GetYourGuide.

// Abstract

Join a team to talk about the journey of GYG with MLOps, from the conception of their platform to the creation of the MLOps engineer role, and to their current stack state.

// Bio

Jean Machado

Jean Carlo Machado is a Data Science Manager at GetYourGuide for the Growth Data Products team and the Machine Learning Platform Team. He is privileged to be able to work on turning ideas in data science from inception to production. Before GYG, Jean was working in a startup in Brazil, building its infrastructure from the ground up. Jean also likes community building and using technology for social good.

Meghana Satish

Meghana Satish is currently working as an MLOps Engineer at GetYourGuide. She has previously held positions at Amazon AWS in Berlin and Microsoft IT in Hyderabad. In addition to her career in technology, Meghana is also a talented singer, dancer, and yoga practitioner.

Olivia Houghton

Olivia has been working as an MLOps engineer at GetYourGuide for the past year and a half or so. Olivia's main work is in building and managing their activity ranking service.

Theodore Meynard

Theodore Meynard, Data Science Manager at GetYourGuide, leads the evolution of their ranking algorithm, enriching customer experiences. His hands-on journey from data scientist to leader has honed his expertise in MLOps and real-time ML. Beyond work, he's a co-organizer of PyData Berlin, underlining his commitment to community and collaborative learning.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

Website: https://investing1012dot0.substack.com/

The Openness of AI report: https://research.contrary.com/reports/the-openness-of-ai

⁠--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Jean on LinkedIn: https://www.linkedin.com/in/jean-carlo-machado-53b15977/

Connect with Meghana on LinkedIn: https://www.linkedin.com/in/meghana-satish-2a825282/?originalSubdomain=de

Connect with Olivia on LinkedIn: https://www.linkedin.com/in/oliviaphoughton/

Connect with Theodore on LinkedIn: https://www.linkedin.com/in/theodore-meynard/

Timestamps:

[00:00] GetYourGuide team's preferred coffee

[00:55] Takeaways

[02:20] Shout out to Berlin MLOps Community

[02:38] Please like, share, and subscribe to our MLOps channels!

[03:39] The GetYourGuide platform

[05:45] GetYourGuide use cases

[11:51] Strong Leadership Vision

[13:59] Creating rituals

[16:55] Feedback on the loop for improvements

[18:35] Different components of GetYourGuide's ML Platform

[21:04] V2 service templates

[24:26] Biggest pain points

[27:02] Feature flags

[30:51] Data foundation

[36:25] Data Testing

[39:53] Cross-team Tool Adoption Process

[44:59] Regrets about design decisions made in the past

[47:53] What's next for the platform with LLMs?

[52:49] Non-data scientists suggesting use cases, language flexibility

[55:14] DevSecOps team's AI study group ideation

[59:25] Experiments in growth data products, marketing split

[1:01:47] Shout out to the Berlin MLOps Community!

[1:03:31] Wrap up

The Centralization of Power in AI // Kyle Harrison // # 181 Oct 13, 2023

Show notes

MLOps podcast #181 with Kyle Harrison, General Partner at Contrary, The Centralization of Power in AI.

// Abstract

Kyle Harrison delves into the limitations imposed by language, underscoring how it can impede our grasp and manipulation of reality while stressing the critical need for improved language model performance for real-time applications. He further explores the perils of centralizing power in AI, with a specific focus on the "Openness of AI", where concerns about privacy are brought to the forefront, prompting his call for businesses to reconsider their reliance on it. The discussion also traverses the evolving landscape of AI, drawing comparisons between prominent machine learning frameworks such as TensorFlow and PyTorch. Notably, the episode underscores the vital role of open-source initiatives within the AI community and highlights the unexpected involvement of Meta in driving open-source development.

// Bio

Kyle Harrison is a General Partner at Contrary, where he leads Series A and growth-stage investing. He joined Contrary from Index, where he was a Partner, and before that, he was a growth investor at Coatue. His portfolio includes iconic startups and public companies, including Ramp, Replit, Cohere, Snowflake, and Databricks. He also regularly shares his analysis on the venture capital landscape via his Substack Investing 101.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related LinksWebsite: https://investing1012dot0.substack.com/

The Openness of AI report: https://research.contrary.com/reports/the-openness-of-ai⁠

--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Kyle on LinkedIn: https://www.linkedin.com/in/kyle-harrison-9274b278/

Timestamps:

[00:00] Kyle's preferred beverage

[00:20] Takeaways

[03:52] Hype in the technology space

[09:20] Application Layer Revenue

[14:44] Stability AI Lawsuit

[18:08] Concern over concentration of power in AI

[20:20] Transparency concerns

[23:35] Open Source AI

[25:57] To use or not to use OpenAI

[30:51] Lack of technical expertise and business-building capabilities

[35:09] AI Transparency and Accountability

[37:50] Traditional ML

[41:47] Finding a unique approach

[45:41] AGI limitations

[47:43] Using Agents

[49:46] Agents getting past demos

[54:39] Tech Challenges & Hoverboard Dreams

[58:04] Both AI hype and skepticism are foolish

[01:27] Wrap up

Adventures in Building CLIP & Other (Largeish) LMs // Sachin Abeywardana // #180 Oct 10, 2023

Show notes

MLOps podcast #180 with Sachin Abeywardana, Deep Learning Engineer at Canva AI, Adventures in Building CLIP and Other (Largeish) Language Models sponsored by Prem AI.

// Abstract

Sachin takes us on an adventure, sharing insights on the pitfalls of not understanding the broader product and the importance of incorporating AI and machine learning capabilities. From the use of AI models for grammar correction and code generation to the fascinating Clip model and the challenges of balancing work and family life, this episode promises to be both informative and thought-provoking.

// Bio

Sachin is the father of two beautiful children. He completed his PhD in Bayesian Machine Learning at the University of Sydney in 2015. In 2016, he discovered Deep Learning and hasn't looked back. He currently works as a Senior Machine Learning Engineer at Canva and is mainly focusing on NLP problems.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

Sachin Blogs: https://sachinruk.github.io/blog.html https://sachinruk.github.io/blog/

Graph ML link: http://web.stanford.edu/class/cs224w/

⁠--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Sachin on LinkedIn: https://www.linkedin.com/in/sachinabeywardana/

Timestamps:

[00:00] Sachin's preferred beverage

[00:26] Takeaways

[02:30] Chat GPT user

[05:58] Understanding of reliable Agents

[08:10] Sachin's background

[12:45] Staying at Deep Learning

[16:17] Recommendation or Lead Scoring

[17:36] Vector database

[19:00] Sachin's blogs

[23:26] The cap people

[26:10] Pursuing a business case

[27:33] Canva

[31:16] Incorporating AI and Machine Learning

[32:17] Sponsor Ad

[38:22] Eliminating unnecessary steps

[39:00] Interacting with the product team

[43:04] Criticisms of the current architecture limitations

[45:58] Insufficient exploration of Transformers

[47:42] Explaining GraphML

[52:35] Fine-tuning ChatGPT2

[57:54] Leading ML Engineers and teams

[59:40] Being practical with Math

[1:05:52] Wrap up

All About Evaluating LLM Applications // Shahul Es // #179 Oct 06, 2023

Show notes

MLOps Coffee Sessions #179 with Shahul Es, All About Evaluating LLM Applications.

// Abstract

Shahul Es, renowned for his expertise in the evaluation space and is the creator of the Ragas Project. Shahul dives deep into the world of evaluation in open source models, sharing insights on debugging, troubleshooting, and the challenges faced when it comes to benchmarks. From the importance of custom data distributions to the role of fine-tuning in enhancing model performance, this episode is packed with valuable information for anyone interested in language models and AI.

// Bio

Shahul is a data science professional with 6+ years of expertise and has worked in data domains from structured, NLP, to Audio processing. He is also a Kaggle GrandMaster and code owner/ ML of the Open-Assistant initiative that released some of the best open-source alternatives to ChatGPT.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

All about evaluating Large language models blog: https://explodinggradients.com/all-about-evaluating-large-language-models

Ragas: https://github.com/explodinggradients/ragas

--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Shahul on LinkedIn: https://www.linkedin.com/in/shahules/

Timestamps:

[00:00] Shahul's preferred coffee

[00:20] Takeaways

[01:46] Please like, share, and subscribe to our MLOps channels!

[02:07] Shahul's definition of Evaluation

[03:27] Evaluation metrics and Benchmarks

[05:46] Gamed leaderboards

[10:13] Best at summarizing long text open-source models

[11:12] Benchmarks

[14:20] Recommending the evaluation process

[17:43] LLMs for other LLMs

[20:40] Debugging failed evaluation models

[24:25] Prompt injection

[27:32] Alignment

[32:45] Open Assist

[35:51] Garbage in, garbage out

[37:00] Ragas

[42:52] Valuable use case besides OpenAI

[45:11] Fine-tuning LLMs

[49:07] Connect with Shahul if you need help with Ragas @Shahules786 on Twitter

[49:58] Wrap up

Building an ML Platform: Insights, Community, and Advocacy // Stephen Batifol // #178 Oct 03, 2023

Show notes

MLOps Coffee Sessions #178 with Stephen Batifol, Building an ML Platform: Insights, Community, and Advocacy.

// Abstract

Discover how Wolt onboards data scientists onto the platform and builds a thriving internal community of users. Stephen's firsthand experiences shed light on the importance of developer relations and how they contribute to making data scientists' lives easier. From top-notch documentation to getting-started guides and tutorials, the internal platform at Wolt prioritizes the needs of its users.

// Bio

From Android developer to Data Scientist to Machine Learning Engineer, Stephen has a wealth of software engineering experience at Wolt. He believes that machine learning has a lot to learn from software engineering best practices and spends his time making ML deployments simple for other engineers. Stephen is also a founding member and organizer of the MLOps.community Meetups in Berlin.

// MLOps Jobs board

jobs.mlops.community

// MLOps Swag/Merch

https://mlops-community.myshopify.com/

// Related Links

⁠--------------- ✌️Connect With Us ✌️ -------------

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

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

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/

Connect with Stephen on LinkedIn: https://www.linkedin.com/in/stephen-batifol/

Timestamps:

[00:00] Stephen's preferred coffee