Context
Red Bull’s venture team sees far more companies than it can study. Flowhive VC was built for that team as an AI workspace for portfolio management and deal flow, and it is still in use.
Business problem
Screening is where a venture team’s attention leaks. Reading every inbound company, checking who else is in the space and keeping a picture of the portfolio all compete for the same hours, and none of it is the actual job, which is deciding where to invest.
My mandate
I own the product: the agents, the workspace around them and its continued operation for the venture team.
Decisions
- Pre-configured agents for the jobs that recur — company due diligence, competitive analysis, market mapping — rather than a general assistant the team has to prompt from scratch each time.
- One workspace holding screening, portfolio intelligence tracking and deal-flow automation, so a company analysis lands next to the portfolio picture it affects.
- The system narrows the field and maps the market; the investment decision stays with the team.
Delivery
The product runs continuously for the venture team and has analysed thousands of startups. I run it.
Outcome
The team screens a volume of companies it could not read by hand, and uses the output to decide where to invest.
Reuse
The agent templates for due diligence and competitive analysis, the market-mapping workflow and the portfolio tracking are product capabilities rather than one-off configurations.
Evidence
The outcome here is qualitative: the volume analysed is described as thousands of startups, and the engagement is ongoing. No investment performance or deal numbers are published.