Lens Your Next Unicorn
Batch-review pitch decks, rank the strongest startups, and give every team a clear report.
Find the leadershidden in your flow.
EvalLens batch-analyzes every deck, compares applications in one format, and brings the strongest candidates to the top.
Launch yourentry point.
Create one place where teams submit their pitch decks. EvalLens collects every application, organizes the full batch, and lets you run evaluation when you're ready.
Share this link with teams — they upload their pitch decks directly.
Application hub
| Team | Materials | Submitted | Status |
|---|---|---|---|
NNovaPay | PDF · repo · video | 2h ago | ReadyProcessingScored |
TTutorVerse | PDF · video | 5h ago | ReadyProcessingScored |
OOrchestrix AI | 1d ago | IncompleteProcessingScored | |
BBrightline | PDF · repo | 2d ago | ReadyProcessingScored |
| + 118 more applications | |||
From messy decks to faster decisions
Instead of manually reading, comparing, and rewriting notes for every deck, your team gets structured reports, batch-level data, and a clear decision trail. This saves reviewer-hours without turning the final choice into a black box.
| Startup | Score | Confidence | Status |
|---|---|---|---|
N NovaPay Cross-border payouts | 8.4 | High | Needs review |
T TutorVerse AI tutoring marketplace | 7.8 | Medium | Open report |
L LedgerLoop Reconciliation for fintechs | 7.2 | High | Reviewed |
O Orchestrix AI Workflow agents for ops | 6.9 | Low | Missing evidence |
From the Newsroom
See allWhat a high-scoring pitch deck actually shows
A field guide to the six dimensions a structured evaluation scores — and why a plain deck that shows its work outranks a beautiful one that doesn't.
ResearchHow do you judge an AI business?
A $2M prize moved the bar from "what can you build" to "does it actually run." The harder truth underneath it — that both AI and human judges are unreliable — is the real reason evaluation, not the model, is the bottleneck.
Press ReleaseWe're building for the Build with Gemini XPRIZE
We're entering the $2M Build with Gemini XPRIZE with a new, AI-operated line of work — and we're going to be honest in public about the bar, because the contest grades the one thing we care about most: evidence.
ResearchSame deck, same score
Reproducibility is the unglamorous property that decides whether a ranking survives scrutiny. Here is the part of the score we made pure math — no model in the loop — so the same deck returns the same number every time.
Press ReleaseFrom AI Jury to EvalLens: what 400+ runs taught us
We thought the answer was more AI judges. After 400+ judging runs, the product taught us the opposite — the work was never the agents. It was the methodology around them, and the human who keeps the final call.
Trusted, Explainable,
Human-Controlled.
Trusted,Explainable,Human-Controlled.
Evidence-based scores
Human in the loop
Methodology
Prompt-injection safety
Private deck vault
Pitch decks are stored and processed inside controlled project workspaces.
See your next cohort ranked in a day
Batch-review every pitch deck, surface the strongest startups, and hand each team an evidence-based report — with the final call always yours.