EvalLens turns decks and application materials into evidence, risks, gaps, and questions your team can use before the meeting. AI prepares the read. Your team decides.
This is the output EvalLens gives your team before the decision meeting — one real pitch report, from deck review to the human vote.
Pitch Report · AI Jury
Pitch deck
AI scores · P1–P6
Project summary
EduPath SEA is an EdTech startup targeting Southeast Asia. The submitted materials show market opportunity and team clearly, but problem, solution, monetization and GTM are weaker.
Strengths
Operator background, complementary team, clear regional category, readable deck.
Weaknesses
No solution slide, no business model, no defined ICP, thin market-sizing evidence.
Pitch report · AI jury
Executive summary
Strong regional market and team signals. The deck still needs a clear business model, customer profile and first-market plan.
Key score signals
Experienced team and an attractive SEA market.
No pricing, revenue model or first-customer path.
Confirm live
Who is the first paying customer, and how does the product make money?
Built first for pitch-deck review. The same first-read logic can support VC dealflow and hackathon application screening — shown in the workflow section below.
You just saw a real report. Send us one batch and we will produce the same from your own decks. Organizations get the first run free through October 31, 2026, on batches of up to 10 decks.
230 submissions.
3 judges.
18 days before finals day.
Everyone is reading different things.
No clean way to explain why team 47 made it and team 89 did not.
Inbound decks came from six sources this week.
The partner meeting is Monday.
Nobody has time to read the full stack.
But missing one strong company still hurts.
Applications are in different formats.
Some have decks. Some have forms. Some have traction claims without proof.
The committee still needs one clean comparison.
Applications close tonight.
The team needs a shortlist before anyone starts building.
Some ideas look strong. Some just look loud.
Manual review does not fail loudly. It fails through uneven depth, scattered notes, and scores that are hard to defend.
AI prepares the read. Your team makes the final decision.
Uploaded materials cannot override the review rules. The rules sit above the content.
Decks and application materials are treated as sensitive by default.
The decision context changes: a finalist board, a partner-ready shortlist, a stronger cohort, or application screening before teams start building.
The core stays the same: structured evidence before the room meets, and a human final call.
Finals are close. Decks are still scattered. Judges are reading at different depth.
An AI panel pre-reads the whole written round on your rubric; your screeners confirm with briefing packs; live rounds run exactly as designed.
University program screening runs on this same workflow.
See the workflowSubmissions just closed. 175 projects, a two hour expo, and four minutes per team per judge.
An AI panel reads every submission on your rubric with execution and technical depth weight protected, and each judge walks the tables with a briefing instead of a blank scorecard.
See the workflowInbound came from six sources. The pipeline meeting is Monday. The median deck gets 150 seconds.
Every submitted deck read in full against your investment dimensions — flags, founder questions, a quote behind every finding. The aide, never the arbiter.
See the workflowThe monthly batch is in. The screening chair is a volunteer. The reading marathon starts tonight.
A one-page, evidence-linked brief per company for your screening committee — findings, red flags, pitch questions. Every judgment stays your members'.
See the workflowHundreds applied in three weeks. The committee decides the cohort next Friday.
Every application read in full and scored on your rubric with page-level evidence. Start with a free retro-test on your last cohort.
See the workflowThe challenge closed with 300+ submissions. The steering committee asks: why these five?
Evidence-linked one-pagers for your jury and a selection memo you can defend upstairs — every pick traced to a quote and a page.
See the workflowReview panels are stretched, cycles run long — and three years later someone asks “why 4.2?”
Every proposal scored to your rubric anchor by anchor, with a record you can produce the day someone asks. Your reviewers remain the reviewers of record.
See the workflowToo slow, the deal lists elsewhere. Too fast, the file won't hold when the NCA asks.
An evidence-linked screening file per project owner, analyst verification, committee decisions logged — built to ESMA's dialect.
See the workflowThe award is announced — and the losing bidder's lawyer asks for the evaluation record.
Every bid read end to end against your published criteria, scored with quotes and page references. Your committee awards; the record is one click.
See the workflowEvalLens prepares the structured first read. Your team makes the final call.
Every workflow on this page leans on the same public methodology — how scores are built, how disagreement is surfaced, and how submissions are protected. Read it before the call, not after.
Tell us what kind of program you run. We will map your workflow, show how the review would work, and walk through a pilot batch.
