Fit Analysis.
Deeper than a PRISM match score. Fit Analysis is TECH's structured partnership assessment — five independently-scored dimensions rolled up to an overall verdict, with a plain-English summary, concrete strengths, specific concerns, dealbreakers, and a recommended approach you can walk into your next partner call with. The tool your investment committee reads before saying yes.
Output: structured 5-dimension score + narrative + strengths + concerns + dealbreakers + recommended approach + exportable PDF.
Latency: 8–15 seconds via Claude Sonnet. PDF export adds another 2–3 seconds.
Usage: typically run 2–3 per week per partnership lead. Complements PRISM (which says "who") with specificity ("why and how").
The five dimensions
Each dimension is independently scored 0–100 against its own rubric. The overall Fit is a weighted combination that emphasises the two we find most predictive of banking-partnership success: Problem Alignment and Regulatory Alignment.
Does your strategic priority for the next 12 months meaningfully intersect with theirs? A "Yes, same thing" scores high; "adjacent but not overlapping" middles; "we're solving different problems" scores low regardless of technical fit.
API surface area, auth model, data format alignment, deployment topology, existing integrations on both sides. We weight more heavily against incompatible auth (e.g., SAML-only partner vs. OAuth-only requester) because it is the #1 source of multi-month integration delays.
KYB posture, data residency, applicable regulations on both sides, consent model, incident history, audit track record. A bank evaluating a fintech partner scores this high when the fintech has passed comparable scrutiny with another tier-1; a fintech evaluating a bank scores high when the bank has previously moved fast with similarly-sized partners.
Decision-making cadence (weekly committee vs. daily standup), risk tolerance, communication style, time zone overlap, engineering culture. The score uses publicly-observable proxies (recent press interviews, career trajectories of senior leaders, reported deal velocity) rather than subjective judgement.
Relative to your five next-best alternatives, does this one move the needle? We benchmark against CORTEX pipeline data — if similar partnerships for similar orgs historically delivered single-digit ACV impact, this dimension scores lower; if this specific target has track record of 20%+ year-one revenue lift for its bank partners, it scores higher.
The narrative sections
Scores tell you how much; the narrative tells you why. Every Fit Analysis produces five written sections in addition to the numeric scorecard.
Summary (3–5 sentences)
A synthesis that executives can take to a meeting without having read the rest. Written to Claude's analytical register — no hype, no hedging beyond what the data supports.
Strengths (3–5 bullets)
Concrete overlap points with evidence. Each bullet cites the source dimension: "Regulatory alignment is strong: partner has three pre-existing tier-1 banking relationships that have passed OCC examinations without material findings." Never the generic "both sides are committed to innovation."
Concerns (2–4 bullets)
Specific risks, each actionable. Not "there might be cultural challenges" — rather, "counterparty moved engineering leadership three times in eighteen months, creating real risk that the technical point-of-contact during integration differs from the one who signed the deal." This is the section investment committees read first.
Dealbreakers (0–2 items)
Hard stops. Usually regulatory (e.g., "partner does not hold the state lending licenses required for your product in TX, CA, NY — any joint go-to-market requires their licensing as a dependency"). If the Dealbreakers section is empty, none were identified.
Recommended approach
How to structure the partnership if you proceed. A specific pilot shape (length, success criteria), a commercial structure (rev share vs. flat fee vs. referral, with suggested splits based on CORTEX benchmarks), a timeline (discovery → pilot → review → contract → launch), and the owners you would want on each side.
How to use it
Fit Analysis has a clear place in the partnership lifecycle:
- After a PRISM match goes from "interesting" to "I'm considering a pilot." Run Fit on the partner org before scheduling a call. Read the Concerns. If they shake your conviction, redirect.
- When an inbound partnership inquiry arrives. Before replying, run Fit on the inbound org. The response you send is informed by what the tool surfaced, not by marketing copy from their deck.
- Before a partnership committee presentation. Attach the PDF export to the committee pack. It is designed to be readable by a bank committee that did not participate in your pre-work.
- Revisit at contract time. Re-run Fit just before signing. If the numbers moved materially, understand why.
Output schema
The API returns structured JSON with every field the UI displays, so you can programmatically ingest Fit Analyses into your partnership CRM, Board reports, or committee packs.
{
"id": "uuid",
"target": { "name": "...", "id": "uuid" },
"overall_fit": 0-100,
"scores": {
"problem_alignment": 0-100,
"technical_compatibility": 0-100,
"regulatory_alignment": 0-100,
"cultural_fit": 0-100,
"strategic_value": 0-100
},
"summary": "3-5 sentences",
"strengths": ["..."],
"concerns": ["..."],
"deal_breakers": ["..."],
"recommended_approach": "structured text",
"ai_meta": {
"model": "claude-sonnet-4-5",
"model_version": "...",
"generated_at": "iso8601"
}
}is_ai_generated: true flag in the API response. Model, version, and prompt template version are preserved in ai_meta for reproducibility — a regulator asking "what did TECH claim about this partnership on 2026-04-18?" can have it answered with receipts.
What Fit does not claim
Fit Analysis is an analytical tool, not a forecasting tool. It does not predict that a partnership will succeed; it systematises and accelerates the analysis a good partnership lead would do over several hours. A high Fit score reduces analysis time; it does not remove the need for judgement, reference calls, or your legal and compliance teams doing their own work. Treat the output as a structured starting point for informed discussion, never a committee substitute.
API surface
Run Fit and read the Concerns first.
The Concerns section is the shortest part of the output and the most valuable. It tells you what to ask about — not what to believe.
Open Intelligence →