Most organisations forget what they decided last quarter — and why. DecisionGraph captures, scores, and connects every decision your team makes.
Every decision moves through capture, maturation, and quality assessment. It is scored continuously by the AI from the moment it enters until it is approved.
You decide what gets captured. DecisionGraph turns it into a structured record — from forwarded emails, pasted text, meeting transcripts, or your own API.
Every signal — an email reply, a transcript, a quiet 36 hours — moves four meters. When all four cross the line, the draft is ready. Until then, it stays a draft.
Approved decisions are scored on the six dimensions of decision quality. The AI walks each one with receipts attached, so you know exactly where the thinking is strong and where it needs another pass.
Frame is well-scoped. Connects pricing to retention metric.
Are we solving the right problem at the right altitude?
Frame is well-scoped. Connects pricing to retention metric.
You can already write decisions down — in a Notion database, a Confluence page, Sharepoint, a slack #decisions channel, etc. Most teams do. But the problem was never storage.
Every decision is graded across six dimensions of decision quality — frame, alternatives, information, trade-offs, reasoning, commitment to action. Drawn from forty years of decision-quality research.A wiki page can’t tell you the thinking is thin on dimension four.
Each decision is embedded and linked into the graph of consequence — what it triggered, what it blocked, what it superseded, what it contradicts. Keyword search finds matching words.We find matching reasoning, eight months and three stakeholder rotations apart.
Stalled drafts, recurring patterns, escalation-of-commitment, sunk-cost drift — surfaced before the next exec meeting, not in it. A custom field stays where you put it.The decisions in here move when something in their world changes.
Put simply, storage answers: “where did we write that down?”
We answer the question of "where did it come from?" and "why was it made?"
If you spend your week chasing what's stalled, defending why you ruled something out in February, or rebuilding the case for a call you already made — DecisionGraph was built around your job.
"We made a call about pricing in February. Then we re-litigated it in May. Then again in August. Nobody can find the original reasoning, the original tradeoffs we considered, or why we ruled out the alternative we're now reconsidering. We're running on institutional amnesia."
Matches D-0612 · D-0884 · 6 stakeholders overlap.
Closely matches D-0612 (Feb '26) and D-0884 (May '26). Three prior threads · 6 stakeholders overlap.
We capture, score, and structure the call — then permanently delete the underlying email or transcript at approval. A cryptographic fingerprint remains, so the provenance is always provable.
Raw text is stored encrypted at rest while the AI runs stateful context merging across the thread. The worker decrypts in-memory only.
When you approve a decision, the raw content payload is permanently purged from our datastore. The decision is recorded but the source vanishes.
Before deletion we store a cryptographic fingerprint of the raw text. Later, the original source from your archive can be matched against it to prove the decision was based on that exact content — without us ever retaining it.
Starting plan for a small team. Generous usage limits
For the leadership unit running a weekly decision cadence across one team.
For multi-team orgs running structured decisions across functions and regions.
Book a 20-minute demo. We'll use one of your real decisions to show you how they map into our tool.