All case studies
Venture Studio / Multi-Entity Holdco · Retainer
A venture studio CEO can now ask questions about his whole portfolio from his phone
A venture studio CEO ran five business lines across dozens of tools and accounts. He didn't want dashboards. We built AI agents that bring the answers to him via iMessage, Slack, and his inbox.
The problem
A venture studio CEO ran oil and gas LPs, SPVs, a food platform, a hemp policy company, and an energy business. Information lived across Airtable, Slack, Google Drive, multiple QuickBooks accounts, and iMessage threads. The team spent hours chasing down basic answers. The CEO did not want dashboards. He wanted to ask a question on his phone and get a plain answer.
What we built
- The CEO can now text questions and get answers. An iMessage assistant does live lookups across contacts and CRM. Slack gets a natural-language agent too: ask "what did we invest in BlueRock at?" and get the answer from the real data.
- Every morning starts with an inbox triage. AI reads his email, flags what matters, and drafts replies in his voice. He reviews and sends instead of writing from scratch.
- Before every meeting, AI writes a one-page brief by pulling context from all entities. No more walking in cold after months without contact.
- A structured data layer keeps every answer accurate. Notion holds the context AI reads. Airtable is the source of truth for people and deals. Supabase is the working data. Without that structure, agents make things up. With it, every answer comes from real records.
- A CRM cleanup sorted thousands of contacts using a new tag system. AI read the freeform notes on every contact and proposed the right tags at scale. That work would have taken a junior analyst months.
- Finance workflows are being added: a unified view across QuickBooks accounts, AI expense categorization, K-1 and 1099 chase agents, and monthly invoice automation.