200 agents do the screening. 5 analysts make the call.
A Fortune 500 Global professional services firm with ~$70B in annual revenue turned weeks of startup research into minutes, giving a 5-person venture team the reach of 200.
Global scale. A venture team of 5, screening 40,000 startups.
A Fortune 500 Global professional services firm. Its corporate venture and corporate development team evaluates startups across every sector the firm operates in.
The market was moving earlier and faster than the team could cover. Screening was the constraint, not judgment.
Where the weeks were going.
Every startup took 2 to 4 weeks to research and score. Against 40,000 new companies a year, the shortfall showed up in 4 places.
Fragmented data
5 to 10 separate databases per research cycle. Analysts spent hours assembling a picture before any evaluation could begin.
Slow evaluation cycles
2 to 4 weeks per startup, with an evaluation framework applied by hand. Delays and inconsistencies were built into the process.
Competitive blind spots
Competitor moves, market shifts and emerging technologies were difficult to track across a scope of 40,000+ startups.
Reporting instead of insight
Analysts spent 30% to 40% of their time building decks and reports rather than interpreting what the data meant.
What they tried first, and where it stopped.
Market databases and research platforms give you data. They do not give you a decision.
The existing tools were static and generic. To take weeks out of a cycle, 4 things had to happen together, on every target.
Reach across every source, not one
Pull from 5 to 10 external databases plus internal datasets, and land the answer in one place.
Apply the same framework every time
Apply the evaluation framework consistently across thousands of companies, instead of interpreting it differently analyst by analyst.
Explain the score, not just produce it
Every recommendation had to be defensible to an investment committee, with the reasoning attached.
Keep watching after the screen
Competitor moves and market shifts had to be monitored continuously, not revisited at the next quarterly review.
A knowledge graph, a team of agents, and one place to ask.
Lyzr designed and implemented an agentic corporate venture platform: a dynamic knowledge graph paired with specialized agents that research, evaluate and monitor startups at scale.
One knowledge graph and an agent layer built on top.
No new research stack to stand up. The platform sits on top of the datasets the team already licenses.
How the platform keeps a score defensible.
Every output has to survive an investment committee.
The team used to gather. Now the team decides.
Before: 10 databases, an evaluation framework and a deck to build, 2 to 4 weeks per startup. After: one request. Agents gather, score and write, and the analyst decides in minutes.
What changed.
Per startup, and across the pipeline.
From one team to the whole ecosystem.
The foundation went live in 12 weeks. The roadmap extends coverage across more industries, adds agents and enhanced scoring models, and opens the platform to partner integrations and API access.
Got a use case in mind?
Platform, engineers and governance all in. We’ll map your workflow against the same 4 tests this one had to pass.
