PRISM Portfolio Analysis

SETA-ED · PRISM Portfolio Analysis

Portfolio Readiness Insight & Scoring Methodology — for funder review teams

For program officers

For each applicant, paste their grant proposal into the PRISM Analysis tab to get a draft scorecard across Prototype, People, and Process readiness plus a populated DARPA-style risk plan. Refine the scores and evidence notes in the Assessment and Risk Plan tabs, compare ventures in Portfolio, and export any scorecard as Markdown for your review team.

PRISM — Portfolio Readiness Insight & Scoring Methodology

See your portfolio's readiness in three dimensions.

PRISM is SETA-ED's portfolio management platform for evaluating ventures across three readiness dimensions: Prototype, People, and Process. Defined by SETA-ED and adapted from the R&D portfolio practices of DARPA, ARPA-E, and other Advanced Research Project Agencies. Paste a grant proposal (or upload a .txt file) to receive a complete draft assessment against all three readiness frameworks — with evidence quotes for every rubric question and an automated risk inventory based on the proposal.

Step 1. Paste the proposal text below. For .docx or .pdf files, open in Word/Preview, select all (⌘/Ctrl+A), and copy-paste here. For .txt files, use the upload button.

Upload .txt: → populates the textarea below

Step 2. Name the venture and run the analysis. The system performs two passes: a readiness scoring pass against all 57 rubric questions, then a risk identification pass that maps proposal content to the DARPA-style risk plan structure.

What the PRISM Analysis returns:

Pass 1 — Readiness scoring. A score (No / Weak / Strong / Yes) for every PRL, PeoML, and ProML question, with a short evidence quote or paraphrase from the proposal supporting each score, and an overall summary of readiness across the three dimensions.
Pass 2 — Risk inventory. A populated DARPA-style risk plan: 4–7 risk categories surfaced from the proposal, each with the venture's stated short-term mitigation, their long-term mitigation plan, and a reviewer-facing confidence rating in the strength of that mitigation.
The whole assessment applied to a new venture in your portfolio — ready for you to review and refine.

Conservative scoring. The analysis is calibrated to score honestly — most ventures overstate their readiness, and the framework's value is in catching that. When the proposal mentions something only briefly or aspirationally, expect "Weak" rather than "Strong." Always re-read the evidence quotes and adjust scores where your judgment differs.

Methodology. The PRISM Analysis is run by a language model executing the SETA-ED rubric. Proposal text is processed in real time and not retained beyond the venture's local record in this browser. The methodology — not the model — is the source of authority: the rubric, scoring anchors, and risk taxonomy were defined by Dr. Melina Uncapher (SETA-ED) drawing on ARPA tiered-support and risk-management practice.

Three-dimensional readiness

Reviewer summary

Export this venture's scorecard for sharing with the review team. Use the branded report for foundation-ready PDF output; use Markdown for drop-in to internal review templates.

Foundation name Logo No branding set — using SETA-ED defaults.

Want help running this across your whole portfolio?

SETA-ED offers methodologist-led deployments — scoring the full portfolio with your team, producing the board-ready and funder-coalition-ready reporting your existing process doesn't generate. Three engagement tiers; the platform itself is included at every tier.

Email SETA-ED Learn more

DARPA-style risk methodology: for each risk, identify the category, decide whether the venture (and you as funder) are willing to accept it, and document explicit short-term and long-term mitigations. Risks without mitigations should be flagged for review or rejected.

Risk category	Status	Short-term mitigation	Long-term mitigation	Confidence

Click any venture name to switch to its assessment. Sort by clicking column headers.

Venture	PRL	PeoML	ProML	Risks logged	Confidence	Updated

Scoring legend. For each rubric question, score the venture's evidence: No no evidence at all · Weak some evidence but not at the level the question asks · Strong clear evidence at the level the question asks · Yes documented evidence with citation/artifact. A level is "achieved" when every question at that level scores Strong or Yes and every level below is achieved.

WriteAhead — worked example

A hypothetical AI-powered writing feedback tool for multilingual learners. Two co-founders, 15 students of pilot data, no formal evidence yet.

Today (engagement start)

PRL

PeoML

ProML

Prototype works in one classroom. Equity stated but not operationalized.

12 months later (target)

PRL

PeoML

ProML

Medium-fidelity prototype validated. Community consultation mechanism documented. Equity-driven design trail in place.

The point for portfolio analysis: ventures whose three readiness scores move together are the credible ones. A venture proposing a leap from PRL 3 to PRL 7 in one cycle (without People and Process advancing) is sandbagging or overstating. Use the rubric to surface this gap before committing funding.

Adds a pre-filled venture so you can explore the rubric with realistic data.