RavenPack

On 4 November, Francisco (Paco) Gómez and I are running a webinar on custom dataset creation with RavenPack Annotations and Bigdata.com. We’ll show how to take an investment hypothesis, turn it into a labelled dataset and get it ready for a backtest. If you have ideas you’ve never had time to test, join us. Register here: https://lnkd.in/eNyViAaN

Investment ideas rarely arrive as data specifications. On November 4, Francisco (Paco) Gómez and I will show how our team turns them into backtest-ready datasets. Most client conversations I have start with a hypothesis: “Companies with real pricing power should protect their margins through this cycle.” “I want to identify which companies are gaining market share in their niche.” The idea is usually sound. The harder part is defining it precisely enough to measure, and then validating that what you measure is actually what you intended. Using pricing power and market share across an actual investment universe, we’ll walk through the process end to end: ▪️ Translating an investment idea into a precise concept definition and structured search plan ▪️ Validating company exposure, removing false positives, and measuring the narrative consistently over time ▪️ Taking the research to scale through exposure heatmaps and backtests, while managing the cost of broader coverage and deeper history Everything is built using RavenPack Annotations and Bigdata.com. The RavenPack Data Science team has been driving the hands-on work behind this research. I’ll cover how we turn the idea into a definition and a search plan, and Paco will take you through validation and the results at scale. Wednesday, November 4, 11:30 AM EST / 5:30 PM CET. 30-minute presentation followed by 15 minutes of Q&A. Register here: https://lnkd.in/eYiWJidR #AI #Alpha #Altdata #investing

Today we launched the Tokenization of Content on Bigdata.com. The insight is simple: the token is already the unit of AI economics. Models consume tokens, inference is billed in tokens. So we tokenized content itself. That means: → Agents retrieve only the tokens that carry the answer, cutting consumption by up to 100× → Every token retrieved is measured, so usage is finally transparent → And because it's measured, it can be priced, and publishers get paid for exactly what agents consume One unit to measure it. One unit to price it. Licensed, cited content from 170+ providers. This is how content economics finally align with AI economics.

Most quant teams consume sentiment. Fewer build it. Running a webinar on RavenPack Annotations: the entity-resolved, knowledge-graph-linked layer that lets you engineer your own features for downstream NLP and ML, rather than inheriting someone else's score. Examples provided using FactSet earnings call transcripts. Register here: https://lnkd.in/ecHVXi2z #Quantinvesting #Alpha #Altdata #webinar #AIinFinance

RBC Capital Markets chose Bigdata.com to power Aiden, their proprietary AI platform and the results speak for themselves. Through Bigdata.com, RBC replaced years of data engineering, vendor management, and manual enrichment with a production-ready financial intelligence layer. Aiden now reasons over timely, market-relevant information in real-time. No hallucinations. Just trusted, grounded, auditable AI at institutional scale. RBC's Head of AI & Digital Transformation, Bobby Grubert, took the stage today at NVIDIA #GTC 2026 to share how RBC is running agentic AI in real production, not as a proof of concept, but delivering measurable ROI. And the numbers tell the story: - Complex client inquiries that took hours (even weeks) now answered in real time - Earnings release analysis cut from 45–120 minutes to seconds - Analyst coverage capacity increased by double digits - Analysis that took 60 minutes per company per quarter, across 1,000+ companies, reduced to mere seconds - Up to $1 billion in AI-driven benefits Bobby is one of the rare AI leaders who has actually shipped this at scale inside a major bank. We're incredibly proud to partner with him and the world-class RBC AI team. On Monday at GTC, Jensen Huang made the case that 90% of the world's data is unstructured and until now, is mostly unsearchable, unqueryable, and essentially invisible to machines. The firms that figure out how to unlock this data for their AI systems won't just have better research. They'll have a structural edge. RBC figured it out early. With bigdata.com. Learn more: https://lnkd.in/gM6begKk

Big milestone for the WorldQuant Brain Competition! On day one, 1,800 builders signed up. A week later, we've hit our goal of 3,000 and the competition is now in full swing. As James Macarthur from Worldquant put it, the vision is to enable people to create new ideas using the Bigdata.com Search API, and build a scalable community of dataset contributors. 3,000 builders. Game on.