Inside Uber’s AI Revolution - Everything about how they use AI/ML

The Rise of Sovereign AI and Global AI Innovation in a World of US Protectionism // MLOps Podcast #331 with Frank Meehan, Founder and CEO of Frontier One AI.Join the Community: https://go.mlops.community/YTJoinInGet the newsletter: https://go.mlops.community/YTNewsletter// Abstract“The awakening of every single country is that they have to control their AI intelligence and not outsource their data" - Jensen Huang. Sovereign AI is rapidly becoming a fundamental national utility, much like defense, energy or telecoms. Nations worldwide recognize that AI sovereignty—having control over your AI infrastructure, data, and models—is essential for economic progress, security, and especially independence - especially when the US is pushing protectionism and trying to prevent global AI innovation. Of course this has the opposite effect - DeepSeek created by a Hedge Fund in China; India building the world's largest AI data centre (3 GW), and global software teams scaling, learning and building faster than ever before. However most countries lack the talent, financing and experience to implement Sovereign AI for their requirements - and it is our belief at Frontier One, that one of the biggest markets for AI applications, cloud services and GPUs will be global governments. We see it already - with $10B of GPUs in 2024 bought directly by governments - and it's rapidly expanding. We will talk about what Sovereign AI is - both infrastructure and software details / why it is crucial for a nation / how to get involved as part of the MLOps community. // BioCo-Founder of Frontier One - building Sovereign AI Factories and Cloud software for global markets.Frank is a 2X CEO | 2X CMO (with 2X exits + 1 IPO NYSE), Board Director (Spotify, Siri) and Investor (SparkLabs Group) with 20+ years of experience in creating and growing leading brands, products and companies.Chair of Improvability, automating due diligence and reporting for corporates, foundations and Governments with AI.Co-founder and partner at SparkLabs Group - investors in OpenAI, Anthropic, 88 Rising, Discord, Animoca, Andela, Vectara, Kneron, Messari, Lifesum + 400 companies in our portfolio. Investment Committee and LP at SparkLabs Cultiv8 with 56 investments in consumer food and regenerative agriculture companies.Co-founder and CMO - later CEO - of Equilibrium AI (Singapore), building it to one of the leading ESG and Carbon data management platforms globally. Equilibrium was acquired by FiscalNote in 2021, where he joined the senior leadership team, running the ESG business globally, and helping the company IPO in 2022 on the NYSE at $1.1B valuation.Board director at Spotify (2009-2012); Siri (2009-2010 exited to Apple); Lifesum (leading AI health app with 50 million users), seed investor in 88 Rising (Asia’s leading independent music label); CEO/CMO and co-founder at INQ Mobile (mobile internet pioneer); and Global Director for devices and products at 3 Mobile.Started as a software developer with Ericsson Mobile in Sweden, after graduating from KTH in Stockholm and the University of Sydney with a Bachelor of Mechanical Engineering, and Master of Science in Fluid Mechanics.// Related Linkshttps://www.frontierone.ai/ and https://www.sparklabsgroup.com~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin 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: /dpbrinkmConnect with Frank on LinkedIn: /frankmeehan/

Thanks to MLflow for supporting this episode — the platform helping teams track, manage, and deploy ML and GenAI projects with ease. Try it free at mlflow.org.What if AI could build and maintain your software—like a co-worker who never forgets state? In this episode, Jiquan Ngiam chats with Demetrios about agents that actually do the work: parsing emails, updating spreadsheets, and reshaping how we design software itself. Less hype, more hands-on AI—tune in for a glimpse at the future of truly personalized computing.// BioJiquan Ngiam is the Co-Founder and CEO of Lutra AI, with deep expertise in artificial intelligence and machine learning. He was previously at Google Brain, Coursera, and in the Stanford CS Ph.D. program advised by Andrew Ng. He helped develop the first online courses in Machine Learning, and is now building agentic AI systems that can complete tasks for us.// Related Linkshttps://www.youtube.com/@LutraAI#api #llm #lutra #costefficiency #latentspace ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Jiquan on LinkedIn: /jngiam/Timestamps:[00:00] Agents That Actually Do Work[08:21] Building Tables With AI Help[12:54] Guardrails for Smarter Code[16:35 - 18:00] MLFlow Ad[18:30] What’s Next for MCP?[23:23] AI as Your Data Conductor[31:13] Rethinking AI + Data Stacks[32:10] Sandbox Security, Real Risks[40:48] Smarter Reviews, Powered by Use[46:08] Cost vs. Quality in AI[52:00] Podcast Editing Gets Creative[56:27] Transparent UIs, Powered by AI[01:00:28] Can AI Learn Good Taste?[01:04:45] Peeking Into Wild AI Futures

Kai Wang joins the MLOps Community podcast LIVE to share how Uber built and scaled its ML platform, Michelangelo. From mission-critical models to tools for both beginners and experts, he walks us through Uber’s AI playbook—and teases plans to open-source parts of it.// BioKai Wang is the product lead of the AI platform team at Uber, overseeing Uber's internal end-to-end ML platform called Michelangelo that powers 100% Uber's business-critical ML use cases.// Related LinksUber GenAI: https://www.uber.com/blog/from-predictive-to-generative-ai/#uber #podcast #ai #machinelearning ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Kai on LinkedIn: /kai-wang-67457318/Timestamps:[00:00] Rethinking AI Beyond ChatGPT[04:01] How Devs Pick Their Tools[08:25] Measuring Dev Speed Smartly[10:14] Predictive Models at Uber[13:11] When ML Strategy Shifts[15:56] Smarter Uber Eats with AI[19:29] Summarizing Feedback with ML[23:27] GenAI That Users Notice[27:19] Inference at Scale: Michelangelo[32:26] Building Uber’s AI Studio[33:50] Faster AI Agents, Less Pain[39:21] Evaluating Models at Uber[42:22] Why Uber Open-Sourced Machanjo[44:32] What Fuels Uber’s AI Team

The Missing Data Stack for Physical AI // MLOps Podcast #328 with Nikolaus West, CEO of Rerun.Join the Community: https://go.mlops.community/YTJoinInGet the newsletter: https://go.mlops.community/YTNewsletter// AbstractNikolaus West, CEO of Rerun, breaks down the challenges and opportunities of physical AI—AI that interacts with the real world. He explains why traditional software falls short in dynamic environments and how visualization, adaptability, and better tooling are key to making robotics and spatial computing more practical.// BioNiko is a second-time founder and software engineer with a computer vision background from Stanford. He’s a fanatic about bringing great computer vision and robotics products to the physical world.// Related LinksWebsite: rerun.io~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin 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: /dpbrinkmConnect with Niko on LinkedIn: /NikolausWestTimestamps:[00:00] Niko's preferred coffee[00:35] Physical AI vs Robotics Debate[04:40] IoT Hype vs Reality[12:16] Physical AI Lifecycle Overview[20:05] AI Constraints in Robotics[23:42] Data Challenges in Robotics[33:37] Open Sourcing AI Tools[39:36] Rerun Platform Integration[40:57] Data Integration for Insights[45:02] Data Pipelines and Quality[49:19] Robotics Design Trade-offs[52:25] Wrap up

LLMs are reshaping the future of data and AI—and ignoring them might just be career malpractice. Yoni Michael and Kostas Pardalis unpack what’s breaking, what’s emerging, and why inference is becoming the new heartbeat of the data pipeline.// BioKostas PardalisKostas is an engineer-turned-entrepreneur with a passion for building products and companies in the data space. He’s currently the co-founder of Typedef. Before that, he worked closely with the creators of Trino at Starburst Data on some exciting projects. Earlier in his career, he was part of the leadership team at Rudderstack, helping the company grow from zero to a successful Series B in under two years. He also founded Blendo in 2014, one of the first cloud-based ELT solutions.Yoni MichaelYoni is the Co-Founder of typedef, a serverless data platform purpose-built to help teams process unstructured text and run LLM inference pipelines at scale. With a deep background in data infrastructure, Yoni has spent over a decade building systems at the intersection of data and AI — including leading infrastructure at Tecton and engineering teams at Salesforce.Yoni is passionate about rethinking how teams extract insight from massive troves of text, transcripts, and documents — and believes the future of analytics depends on bridging traditional data pipelines with modern AI workflows. At Typedef, he’s working to make that future accessible to every team, without the complexity of managing infrastructure.// Related LinksWebsite: https://www.typedef.aihttps://techontherocks.showhttps://www.cpard.xyz~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Kostas on LinkedIn: /kostaspardalis/Connect with Yoni on LinkedIn: /yonimichael/Timestamps:[00:00] Breaking Tools, Evolving Data Workloads[06:35] Building Truly Great Data Teams[10:49] Making Data Platforms Actually Useful[18:54] Scaling AI with Native Integration[24:04] Empowering Employees to Build Agents[28:17] Rise of the AI Sherpa[36:09] Real AI Infrastructure Pain Points[38:05] Fixing Gaps Between Data, AI[46:04] Smarter Decisions Through Better Data[50:18] LLMs as Human-Machine Interfaces[53:40] Why Summarization Still Falls Short[01:01:15] Smarter Chunking, Fixing Text Issues[01:09:08] Evaluating AI with Canary Pipelines[01:11:46] Finding Use Cases That Matter[01:17:38] Cutting Costs, Keeping AI Quality[01:25:15] Aligning MLOps to Business Outcomes[01:29:44] Communities Thrive on Cross-Pollination[01:34:56] Evaluation Tools Quietly Consolidating

What makes a good AI benchmark? Greg Kamradt joins Demetrios to break it down—from human-easy, AI-hard puzzles to wild new games that test how fast models can truly learn. They talk hidden datasets, compute tradeoffs, and why benchmarks might be our best bet for tracking progress toward AGI. It’s nerdy, strategic, and surprisingly philosophical.// BioGreg has mentored thousands of developers and founders, empowering them to build AI-centric applications.By crafting tutorial-based content, Greg aims to guide everyone from seasoned builders to ambitious indie hackers.Greg partners with companies during their product launches, feature enhancements, and funding rounds. His objective is to cultivate not just awareness, but also a practical understanding of how to optimally utilize a company's tools.He previously led Growth @ Salesforce for Sales & Service Clouds in addition to being early on at Digits, a FinTech Series-C company.// Related LinksWebsite: https://gregkamradt.com/YouTube channel: https://www.youtube.com/@DataIndependent~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Greg on LinkedIn: /gregkamradt/Timestamps:[00:00] Human-Easy, AI-Hard[05:25] When the Model Shocks Everyone[06:39] “Let’s Circle Back on That Benchmark…”[09:50] Want Better AI? Pay the Compute Bill[14:10] Can We Define Intelligence by How Fast You Learn?[16:42] Still Waiting on That Algorithmic Breakthrough[20:00] LangChain Was Just the Beginning[24:23] Start With Humans, End With AGI[29:01] What If Reality’s Just... What It Seems?[32:21] AI Needs Fewer Vibes, More Predictions[36:02] Defining Intelligence (No Pressure)[36:41] AI Building AI? Yep, We're Going There[40:13] Open Source vs. Prize Money Drama[43:05] Architecting the ARC Challenge[46:38] Agent 57 and the Atari Gauntlet

Bridging the Gap Between AI and Business Data // MLOps Podcast #325 with Deepti Srivastava, Founder and CEO at Snow Leopard.Join the Community: https://go.mlops.community/YTJoinInGet the newsletter: https://go.mlops.community/YTNewsletter// AbstractI’m sure the MLOps community is probably aware – it's tough to make AI work in enterprises for many reasons, from data silos, data privacy and security concerns, to going from POCs to production applications. But one of the biggest challenges facing businesses today, that I particularly care about, is how to unlock the true potential of AI by leveraging a company’s operational business data. At Snow Leopard, we aim to bridge the gap between AI systems and critical business data that is locked away in databases, data warehouses, and other API-based systems, so enterprises can use live business data from any data source – whether it's database, warehouse, or APIs – in real time and on demand, natively. In this interview, I'd like to cover Snow Leopard’s intelligent data retrieval approach that can leverage business data directly and on-demand to make AI work.// BioDeepti is the founder and CEO of Snow Leopard AI, a platform that helps teams build AI apps using their live business data, on-demand. She has nearly 2 decades of experience in data platforms and infrastructure.As Head of Product at Observable, Deepti led the 0→1 product and GTM strategy in the crowded data analytics market. Before that, Deepti was the founding PM for Google Spanner, growing it to thousands of internal customers (Ads, PlayStore, Gmail, etc.), before launching it externally as a seminal cloud database service. Deepti started her career as a distributed systems engineer in the RAC database kernel at Oracle.// Related LinksWebsite: https://www.snowleopard.ai/AI SQL Data Analyst // Donné Stevenson - https://youtu.be/hwgoNmyCGhQ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin 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: /dpbrinkmConnect with Deepti on LinkedIn: /thedeepti/Timestamps:[00:00] Deepti's preferred coffee[00:49] MLflow vs Kubeflow Debate[04:58] GenAI Data Integration Challenges[09:02] GenAI Sidecar Spicy Takes[14:07] Troubleshooting LLM Hallucinations[19:03] AI Overengineering and Hype[25:06] Self-Serve Analytics Governance[33:29] Dashboards vs Data Quality[37:06] Agent Database Context Control[43:00] LLM as Orchestrator[47:34] Tool Call Ownership Clarification[51:45] MCP Server Challenges[56:52] Wrap up

The Creator of FastAPI’s Next Chapter // MLOps Podcast #324 with Sebastián Ramírez, Developer at FastAPI Labs.Join the Community: https://go.mlops.community/YTJoinInGet the newsletter: https://go.mlops.community/YTNewsletter// AbstractThe creator of FastAPI is back with a new chapter—FastAPI Cloud. From building one of the most loved dev tools to launching a company, Sebastián Ramírez shares how open source, developer experience, and a dash of humor are shaping the future of APIs.// BioSebastián Ramírez (also known as Tiangolo) is the creator of FastAPI, Typer, SQLModel, Asyncer, and several other widely used open-source tools. He has collaborated with companies and teams around the world—from Latin America to the Middle East, Europe, and the United States—building a range of products and custom solutions focused on APIs, data processing, distributed systems, and machine learning. Today, he works full time on FastAPI and its growing ecosystem.// Related LinksWebsite: https://tiangolo.com/FastAPI: https://fastapi.tiangolo.com/FastAPI Cloud: https://fastapicloud.com/FastAPI for Machine Learning // Sebastián Ramírez // MLOps Coffee Sessions #96 - https://youtu.be/NpvRhZnkEFg~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin 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: /dpbrinkmConnect with Tiangolo on LinkedIn: /tiangoloTimestamps:[00:00] Sebastián's preferred coffee[00:15] Takeaways[01:43] Why Pydantic is Awesome[06:47] ML Background and FastAPI[10:44] NASA FastAPI Emojis[15:21] FastAPI Cloud Journey[26:07] FastAPI Cloud Open-Source Balance[31:45] Basecamp Design Philosophy[35:30] AI Abstraction Strategies[42:56] Engineering vs Developer Experience[51:40] Dogfooding and Docs Strategy[59:44] Code Simplicity and Trust[1:04:26] Scaling Without Losing Vision[1:08:20] FastAPI Cloud Signup[1:09:23] Wrap up

Willem Pienaar and Shreya Shankar discuss the challenge of evaluating agents in production where "ground truth" is ambiguous and subjective user feedback isn't enough to improve performance.The discussion breaks down the three "gulfs" of human-AI interaction—Specification, Generalization, and Comprehension—and their impact on agent success.Willem and Shreya cover the necessity of moving the human "out of the loop" for feedback, creating faster learning cycles through implicit signals rather than direct, manual review.The conversation details practical evaluation techniques, including analyzing task failures with heat maps and the trade-offs of using simulated environments for testing.Willem and Shreya address the reality of a "performance ceiling" for AI and the importance of categorizing problems your agent can, can learn to, or will likely never be able to solve.// BioShreya ShankarPhD student in data management for machine learning.Willem PienaarWillem Pienaar, CTO of Cleric, is a builder with a focus on LLM agents, MLOps, and open source tooling. He is the creator of Feast, an open source feature store, and contributed to the creation of both the feature store and MLOps categories. Before starting Cleric, Willem led the open source engineering team at Tecton and established the ML platform team at Gojek, where he built high scale ML systems for the Southeast Asian decacorn.// Related Linkshttps://www.google.com/about/careers/applications/?utm_campaign=profilepage&utm_medium=profilepage&utm_source=linkedin&src=Online/LinkedIn/linkedin_pagehttps://cleric.ai/~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Shreya on LinkedIn: /shrshnkConnect with Willem on LinkedIn: /willempienaarTimestamps:[00:00] Trust Issues in AI Data[04:49] Cloud Clarity Meets Retrieval[09:37] Why Fast AI Is Hard[11:10] Fixing AI Communication Gaps[14:53] Smarter Feedback for Prompts[19:23] Creativity Through Data Exploration[23:46] Helping Engineers Solve Faster[26:03] The Three Gaps in AI[28:08] Alerts Without the Noise[33:22] Custom vs General AI[34:14] Sharpening Agent Skills[40:01] Catching Repeat Failures[43:38] Rise of Self-Healing Software[44:12] The Chaos of Monitoring AI

Tricks to Fine Tuning // MLOps Podcast #318 with Prithviraj Ammanabrolu, Research Scientist at Databricks. Join the Community: https://go.mlops.community/YTJoinIn Get the newsletter: https://go.mlops.community/YTNewsletter // AbstractPrithviraj Ammanabrolu drops by to break down Tao fine-tuning—a clever way to train models without labeled data. Using reinforcement learning and synthetic data, Tao teaches models to evaluate and improve themselves. Raj explains how this works, where it shines (think small models punching above their weight), and why it could be a game-changer for efficient deployment.// BioRaj is an Assistant Professor of Computer Science at the University of California, San Diego, leading the PEARLS Lab in the Department of Computer Science and Engineering (CSE). He is also a Research Scientist at Mosaic AI, Databricks, where his team is actively recruiting research scientists and engineers with expertise in reinforcement learning and distributed systems.Previously, he was part of the Mosaic team at the Allen Institute for AI. He earned his PhD in Computer Science from the School of Interactive Computing at Georgia Tech, advised by Professor Mark Riedl in the Entertainment Intelligence Lab.// Related LinksWebsite: https://www.databricks.com/~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin 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: /dpbrinkmConnect with Raj on LinkedIn: /rajammanabroluTimestamps:[00:00] Raj's preferred coffee[00:36] Takeaways[01:02] Tao Naming Decision[04:19] No Labels Machine Learning[08:09] Tao and TAO breakdown[13:20] Reward Model Fine-Tuning[18:15] Training vs Inference Compute[22:32] Retraining and Model Drift[29:06] Prompt Tuning vs Fine-Tuning[34:32] Small Model Optimization Strategies[37:10] Small Model Potential[43:08] Fine-tuning Model Differences[46:02] Mistral Model Freedom[53:46] Wrap up