🧠

Thousands of line items, scattered filings, and hours of stitching data. We’ve expanded our capabilities to ETFs.

Banewick is here

@cf0_ai product update: Every AI agent startup is using RAG pipelines. We never built it. Most AI finance platforms chunk documents, embed them in a vector database, and hope the model retrieves the right fragment. We took a different approach. cf0 ingests regulatory filings from 11 countries, parses them into clean structured sections, and mounts them directly into an isolated sandbox for the AI agent. No embeddings. No retrieval scores. Just files. The agent reads filings the way an analyst reads filings. It opens the MD&A. It greps across three years of risk factors. It pulls line items from the financial statements to build a valuation. It compares segment data across competitors. Every answer traces back to a specific section of a specific filing. You can open the same source and verify it yourself. Each user's sandbox has four mount points: /private : your personal workspace. Artifacts, notes, saved analyses. Read and write. Persists across sessions. /shared : your organisation's shared data. Documents and research your team has uploaded. Read only. /public : platform-wide skills and research templates. The institutional knowledge that powers every analysis. /data/filings : every regulatory filing we've ingested across 11 markets. Parsed, sectioned, ready to read. Credentials are scoped per session. User A physically cannot access User B's data. The infrastructure enforces it. Here's why this architecture matters: RAG is scaffolding. It exists because models can't process full documents yet. But models are getting better fast. Teams that built everything around chunking and retrieval will spend ages ripping it out. We never built it. Our proprietary data pipeline normalizes filings from the SEC, EDINET, BSE, and other regulatory bodies into a universal format. One pipeline. Any country. Any language. When a better model ships, it reads the same files and produces deeper analysis. We change nothing. Clean data in. Better model. Better output. The best infrastructure is the kind that improves without you touching it. Back to work.

@cf0_ai product update: We just made every public filing on earth searchable in under a second. Here's the problem with financial research today: if you're an analyst at a hedge fund or family office, getting access to regulatory filings outside the US is painful. You're navigating Japanese regulatory portals in a language you don't read. You're waiting for someone on your team to manually download PDFs from the Indian stock exchange. You're losing hours before you even start analyzing. Unacceptable. cf0 now indexes 28,000+ public companies across 11 markets — United States, Japan, India, Australia, South Korea, UK, Canada, and more. You search a company. You see every filing they've ever made. You click "Ingest". The filing downloads, gets parsed into clean sections, and is ready for our AI to analyze in seconds exactly like an analyst would. The part I'm most proud of: we don't need custom parsers for each country. One universal pipeline handles everything — a 200-page Japanese annual securities report and a US 10-K go through the exact same system. Any language. Any format. Any regulatory body. This is what we believe financial infrastructure should look like. Not walled gardens. Not manual processes. Not US-only. Back to work.

Body obeys mind. No sports. No running. No weights. Red meat, coffee, nicotine. Brain definitely making up for 45% of my BMR. Back to work.

Absolutely fucking massive update coming to cf0

Over the past 6 months I have discovered that I have incredible psychological stability

Some didn’t like the idea of a changelog being represented through a knowledge graph. These are firms / investors I don’t want to work with. Clearly they have no ability to see where the world is heading :)

Excited to develop this forward with all of our relevant updates. Still much to work out on it but we have so many interesting applications planned for it both internally and with firms.

Quick overnight redesign

I am fully diverting my energy into making every aspect of our company as AI native as possible

cf0.ai has officially signed a deal with FINEDA to power their analysts with our AI financial research platform, improving the quality, efficiency, and depth of financial analysis across their workflows while significantly reducing operational costs. This means that repetitive tasks such as data aggregation, analysis, model building, and audit processes will be supported by cf0.ai, allowing FINEDA’s analysts to focus their time and expertise on the insights and decisions that truly matter, while also integrating their proprietary methodologies directly into the workflow through @cf0_ai Skills.

Bloomberg terminal won’t be replaced.

We indexed every SEC filing. Section by section. Item by item. Searchable in milliseconds. @cf0_ai navigates filings the way you do. And you can teach it your exact workflow for any analysis. Your process, on every company, at machine speed. Seconds, not hours. cf0.ai

US military is just too cool.

Most X articles I see from founders are just bland and hyperbolic iterations of the most obvious crap lol Then I see people on my feed saying “must read if you’re a founder”. So many in the industry are just mindless drones with no intellectual capacity to produce new ideas or conceptualize where the market is / where it’s going.

Gn

10/ Join the waitlist if you want to try the interface yourself. No hype. Just a different way to structure research. → app.cf0.ai/experience

9/ We’re opening an invite-only waitlist for analysts doing: – research – reporting – ongoing coverage If that’s your day-to-day, this should feel familiar.

Financial research has outgrown its tools. cf0 is what comes next. We’re opening a waitlist for analysts, investors and PMs → app.cf0.ai/experience 🧵