For funds
Diligence you can put in the memo — and defend in the committee.
Make faster, safer calls with cited, auditable diligence in hours — white-labelled under your brand, run in your own cloud, and called straight from your deal pipeline. When a partner challenges a finding, you point to its public source and an audit hash.
White-label, dedicated-cluster, volume API.
The problem
A missed red flag shouldn't surface after the wire.
Diligence done the old way is fine until a sanctions hit or a lawsuit turns up post-investment. You need a read fast enough to inform the deal, and cited well enough to survive your committee's scrutiny.
What you get
Built for how a fund actually screens and decides.
White-label scorecards
Every cited scorecard, audit trail, and export carries your brand — not ours. Put it in front of your investment committee and LPs as your own work product.
Dedicated-cluster mode
Run the full data path in your own AWS or Azure account. Your subjects, your tenancy, your keys — the public control plane is never in the auth path.
Volume API & MCP
Call diligence straight from your deal pipeline. Submit a whole portfolio of companies in one batch and pull cited scorecards back programmatically.
Per-deal isolation
Each deal is partitioned. Runs, citations, and audit trails stay separated, with a runtime guard that refuses cross-deal reads.
Audit export
One click pulls a tamper-evident, hash-chained audit bundle per company — the evidence your IC and compliance team need to trust the score.
Hours, not days
A multi-day, analyst-heavy diligence cycle becomes a same-day cited scorecard — fast enough to build into the deal, not after it has moved on.
Diligence workflow
Company in, defensible scorecard in the memo.
Drop in the company or founder
Enter a company or founder — a name and any identifiers (domain, registration number, LEI) — or POST it to the API. The engine resolves the entity and its officers automatically.
Ground every finding in public data
Adverse media, sanctions, PEP, litigation, officers, funding, hiring, and patents are computed against SEC EDGAR, Companies House, OpenCorporates, sanctions lists, CourtListener, GDELT, and USPTO — each finding cited to its source record.
Defend the score in the memo
Drop the cited risk × upside scorecard into your IC memo under your brand. When a partner challenges a litigation flag or a sanctions hit, point to the citation and the audit hash.
For your ops & compliance team
The questions your LPs ask first.
Where does our data live, and who can see it?
In dedicated-cluster mode the entire data path runs in your own cloud account. The public control plane is not in your auth path and cannot read tenant data — that boundary is enforced at the VPC and again at the data-tier client.
How are deals kept separate?
Each deal is partitioned by tenant ID, enforced at the query layer and re-checked by a runtime isolation guard that raises on any cross-tenant access. Dedicated clusters add a full account-level boundary on top.
Do our subjects ever reach a third-party model?
No. Inference runs on a self-hosted Qwen fleet on our own (or your own) GPUs. Subjects you assess are never sent to a third-party API, and we never train a model on your data.
What can we hand to our IC or an auditor?
A per-company, hash-chained audit bundle that records every action, actor, timestamp, payload hash, and reason code — exportable and anchored for tamper-evidence, so any finding can be independently re-verified.
What is the compliance posture today?
The tamper-evident audit chain, GDPR design posture, and per-tenant isolation are implemented. SOC 2 Type II is in progress — see the Trust page for the current, honestly-labelled status of each control.
Volume API & batch submission
Decision-support only — not investment advice
Put a cited scorecard in your next memo.
Book a demo and we'll run a real company through the platform — white-labelled, cited, and on a tamper-evident audit trail you can defend in front of your committee.