Data Council

This month, Pedram Navid, Chief Dashboard Officer at Dagster Labs, curates ten essential reads from the world of data & AI. Read here on LinkedIn or on the Data Council blog at: https://lnkd.in/gTQvu_f7 --- Articles written by: Hannes Mühleisen Mark Raasveldt Amanda Fioritto Jacob Matson Alexandre Salama Ellese Cotterill Pedram Navid Tim Tully Joff Redfern Derek Xiao Aliaa Abbas Anupriti Warade, MBA, PMP® Jacob T.

Great episode! Appreciate how deep Spencer Kimball went on how Cockroach Labs incorporated open source and open core licensing to keep up with customer needs.

Your December dose of data & AI is here! In this month's roundup, you'll get: - Our Data Council 2025 talk tracks (save $ and get Early Bird tickets now) - Feature on LlamaIndex an innovative project for building RAG apps - A new podcast featuring Spencer Kimball, exploring how Cockroach Labs open source licensing to grow their business - Curated technical reads and select job listings from the data & AI community Happy reading!

What a year for data infrastructure and AI engineering. And a defining milestone for Zero Prime Ventures in 2024: closing our $48M Fund II to double down on backing Day 0 engineer-founders building foundational infrastructure for data & AI. Here's a look back at 2024. A few highlights: → Data Council's biggest year yet—technical talks from Meta, Berkeley, Anthropic showcasing how quality data + smart architecture beat massive budgets → Portfolio companies raised significant rounds (Freshpaint's $30M Series B, Orb's $25M Series B, CuspAI's $30M Seed) → Zero Prime Ventures Podcast hit 2 years of deep technical conversations with top engineer-founders Three posts that resonated with our community: 1. Berkeley's AI Architecture Breakthrough - Student team built competitive model for $1,000 in three days - "Better architecture, cleaner data, and smarter engineering beat massive compute budgets every time." https://lnkd.in/g3wnNXvn 2. Knowledge Graph Renaissance with AI - Unique semantic capabilities for reducing hallucinations - Infrastructure opportunity beyond vector databases https://lnkd.in/gZuRv7pr 3. Data Council is Heading Home - 2025 Data Council will be bringing together the brightest minds in data as it returns to the Bay Area - Technical depth remains our north star https://lnkd.in/gC_Jd5cg Deeply grateful to our community of engineers and technical founders who've trusted us as partners in their journey from Day 0. Looking ahead: Data Council returns to SF Bay Area (April 22-24, 2025). 3 days of cutting-edge technical talks and real-life insights from the brightest minds in data & AI. Hope to see you there! To every engineer-founder laying the foundation for AI's future: thank you for an incredible 2024.

A Berkeley student team built a competitive AI model for $1,000 in three days. No massive compute farm. No billions in funding. Just smart engineering and quality data. Their results challenge our assumptions about AI development. The team used 700MB of carefully curated, human-validated conversations—and outperformed massive datasets backed by billions in compute. Let that sink in. The real breakthrough was their architecture. Through smart design choices they reduced compute requirements so dramatically that models could run on standard laptops. Raw compute power matters far less than we thought. One of my highlights from Data Council '24 was watching former U.S. Chief Data Scientist DJ Patil and Joseph Gonzalez dive deep into these findings. Joey, who leads Berkeley's RISELab, laid out the evidence clearly: targeted models consistently outperform general ones in specific domains. This reinforces a core belief about AI development: Better architecture, cleaner data, and smarter engineering beat massive compute budgets every time. Data is the mother of AI! Watch the full conversation "Why it Takes Billions: Navigating the AI landscape w/ OpenAI, Google and Nvidia" in comments.

Your December dose of data & AI is here! In this month's roundup, you'll get: - Our Data Council 2025 talk tracks (save $ and get Early Bird tickets now) - Feature on LlamaIndex an innovative project for building RAG apps - A new podcast featuring Spencer Kimball, exploring how Cockroach Labs used open source licensing to grow their business - Curated technical reads and select job listings from the data & AI community Happy reading! https://lnkd.in/g82mntTM --- Reads written by: Mosbeh Barhoumi, Animesh Kumar, Eric Xiao, Adel Zaalouk, Simon Späti, Mehdi Ouazza, Julien Hurault, Kevin Liu, LangChain, Siddhant Sahu, Nikos Kafritsas, Adam Bellemare Jobs at: MotherDuck Clusterfudge Hex Dagster Labs Soda Vectara

Introducing our Data Council 2025 Track Hosts! We’ve assembled the sharpest technical minds to curate your learning experience: —-- Shawn swyx W (Smol AI): has built dev tools and communities at AWS, Netlify, Temporal and Airbyte Daniel Francisco (Meta): Shipped AI products to 100M+ users. Ex-Head of Product (Search) at Google Sean Taylor (Motif): built data science teams at Lyft and Meta where he cracked some of tech's hardest causality problems, PhD-level stats expertise Tristan Zajonc (Continual): Scaled ML at Cloudera post Sense acquisition. Spent a decade applying data science at World Bank and Harvard (PhD) before jumping into AI Maggie Hays (DataHub): Community PM for DataHub and Founding Team at Acryl Data Carlos Aguilar (Hashboard): Built warehouse-scale robotics at Amazon/Kiva & ML + robotics at Cornell’s Computational Synthesis Lab Sai Krishna Srirampur (Clickhouse): built Postgres tools at Microsoft, early Citus Data employee (acquired by Microsoft) Scott Breitenother (Brooklyn Data): investor and advisor who specializes in building data-driven orgs Sean Anderson (Vectara): Veteran of Rackspace, Cloudera and StreamSets —-- Each track host is curating technical deep-dives from speakers who are building and scaling real systems. No fluff, no marketing talks—just pure technical content from the trenches. Join us April 22-24 in the SF Bay Area for Data Council 2025 to learn how top teams are building and scaling data/AI systems. Early bird tickets now live 🎟️

📚 New Blog: Must-read data articles for November 2024's Top Ten This month, Lindsay MacDonald from Monte Carlo asks a critical question: Is data ready for GenAI? While AI seems ready to take off, are our data foundations prepared? Let’s find out: https://lnkd.in/gNNkTk59 Mission Lane: Built an always-on AI compliance testing system for fintech Confluent | Adam Bellemare: How to handle bad data in event streams Yelp: Optimizing Redshift data loading with DBT & Spectrum Fowler Martin | Kiran Prakash: Using fitness functions to scale data product governance Mikkel Dengsøe: How 40 top data teams structure their orgs Felicis | Astasia Myers Eric Flaningam: AI data infrastructure landscape and future bets Jack Vanlightly: Balancing incremental processing vs data quality Valliappa Lakshmanan: Why your architecture needs a "platinum" layer Barr Moses: Survey: 91% building AI but 2/3 don't trust their data 🚨 Ben Lorica 罗瑞卡: How leading AI labs prioritize data quality first Want to curate a monthly top ten? Contact us at: community@datacouncil.ai

Last Call! Our Call for Proposals for Data Council 2025 ends tomorrow. We’re searching for the sharpest technical minds in data engineering, MLOps, AI and infrastructure to share their insights. Why submit? → Showcase your work to a global audience of top-tier engineers and founders → Network with talent from companies like Meta, Vectara and Anthropic → Get feedback on your ideas from industry veterans → Potentially catch the eye of investors (like me 😙) If you’re building something groundbreaking, we want to hear about it! Apply here: https://lnkd.in/gvEnaXPt Or if you know someone else who is, tag them in a comment so they see this post. 🗓 See you April 22-24 in the SF Bay Area.

The current cycle of AI tooling focuses too heavily on primitive chat use cases. The next generation will be fundamentally different—requiring capabilities that mirror how people actually work: → Long-term planning abilities and multi-step reasoning to pursue complex goals → Persistent memory and knowledge management to learn from mistakes and develop skills → Tool-building capabilities to create and optimize solutions This shift from chat to semi-autonomous agent assistants represents a massive opportunity in AI infra. LLM apps are everywhere. The hard technical problems—and bigger opportunities—are in designing systems, knowledge bases, backend data workflows and tool automation. This is what powers serious enterprise AI. Ashwin Ramesh's talk at Data Council 2024 validated this thesis: AI's role is transforming from conversation partner to technical collaborator and autonomous agent. The infrastructure to enable this is being built now. If you're building AI infra or devtools, this talk is worth 15 minutes of your time. Ashwin digs into specific research projects tackling planning systems, memory retention and tool creation. Link to talk in comments.