EvalLense walks you from sign-in to a ranked leaderboard. The path is fixed, and each step lights up as you scroll.
01Sign inEmail and password, or Google OAuth. You land in your own workspace.
02DashboardYour projects and their status, with a place to start a new one.
03Set up the projectA five-step wizard: details, criteria and weights, judges, how decks are collected, and a final review. Mode — Pitch Competition — is chosen before the wizard opens.
04Collect decksAdd applications by hand, or share a self-upload link and let teams submit their own.
05Start judgingThe AI pipeline runs across every deck in the batch.
06Jury reviewYou set the Jury Score on each dimension, guided by the AI report.
07LeaderboardRanked by your Final Score, not by the AI score.
Image · organizer flow · 4:3A 7-step track, nodes lighting in sequence on an ink surface
Evaluation pipeline
Every deck runs the samefive stages
Each application passes a fixed pipeline. The numeric layer is deterministic — the same findings and weights produce the same AI Total Score, an advisory reference. Ranking is built from your score, not this one.
01
Decoder
Brings any deck — PDF, PPTX, or Google Slides — to a structured, slide-level view.
02
AI Judges
Six independent judges score the deck across P1–P6, in parallel.
03
Summarizer
Function 1 runs deterministic math; Function 2 writes the narrative and the questions to ask each team.
04
Scoring
Your criterion weights apply to the human Jury Score to form the Final Score.
05
Report
An explainable report is assembled for every participant.
Three modules
One product, three connected modules
An Entry Hub for intake, Evidence-Based Reports for the analysis, and a Review Board where a person makes the call.
Entry Hub
One entry point for the batch
Applications and decks land in a single place — added by hand or through a self-upload link, ready to evaluate.
Image · modules bento · 16:9Three module tiles with quiet signal icons, one lens accent
Evidence-Based Reports
Scores tied to evidence
Every score and conclusion links back to evidence in the deck, not to a black-box verdict.
Review Board
The board for the human decision
Review, compare, and rank — with a leaderboard built from your Final Score.
AI jury
six lenses Six independent judges, six lenses
Evaluation runs through a jury of six AI judges, each with its own lens across P1–P6. They work independently and never see one another's scores — and where they disagree, the report shows it.
J-P1
Problem
The pain, the user, the urgency, and the alternatives a deck claims to beat.
J-P2
Solution Logic
Product logic, differentiation, and how coherently the solution holds together.
J-P3
Business Value / Market
The market, the value, and how the business intends to make money.
J-P4
Pitch Quality
Clarity, narrative, structure, and delivery.
J-P5
Team Readiness
Founder-market fit, skills, and the ability to execute.
J-P6
Feasibility
Roadmap, resources, and operational realism.
Why it matters
Less manual review, a clearer trail
EvalLense cuts the hours spent reading decks by hand, holds every evaluation to one standard, and leaves your team a decision trail it can defend.
Six AI lenses do the reading. You make the decision — with the evidence in front of you.
Pitch Competition · P1–P6
What you get
Everything a run leaves behind
After a batch runs, you're left with a structured set of outputs — not a folder of scattered files and threads.
Workspace
Structured workspace
Every application in one organized space, instead of scattered files and threads.
Evaluations
Deck-level evaluations
Scores across P1–P6 with a confidence signal on each.
Evidence
Judge assessments & evidence
Each assessment tied back to evidence in the deck.
Signals
Strengths, weaknesses, gaps
Surfaced alongside deck-completeness signals.
Leaderboard
Leaderboard & comparison
Ranked by your Final Score.
Get started
See the whole workflow on your own decks
Book a demo and watch intake, evaluation, and human review play out end to end.
