Process graphs, not knowledge graphs.
Every software company between 50 and 500 people has the same problem. Knowledge is scattered across ten tools. Decisions made in Slack disappear within days. New hires take 3–6 months to ramp. Senior people leaving means losing tribal knowledge that no system ever captured.
Enterprise-search products solved this for the Fortune 500, at a $50K/year floor. Pulse solves it for the rest of us — with a different bet.
Three structural choices, all interlocked.
First, we model process, not just knowledge. Features, decisions, and commitments are first-class entities. That’s what enables Feature Archeology, Decision Memory, and accurate effort estimation. A pure document index can’t express “the auth feature took 3 weeks longer because of OAuth complications.” Ours does.
Second, onboarding is a 30-minute interview, not a 4-week IT engagement. We talk to the founder. We learn the vocabulary. The system gets a working model of your company before connectors come online. Permission-aware AI shouldn’t require IT.
Third, we charge by verified outcomes, not by seats. $19 / $49 / $99 per seat plus $5 per agent action. Aligned to value, hostile to the legacy per-seat enterprise economics that keep mid-market teams locked out.
AI that does, not just suggests.
Most AI products stop at a chat reply. Pulse drafts the actual action — the Slack DM to your stalled PR’s owner, the Linear ticket from a captured decision, the calendar invite for tomorrow’s kickoff — and waits for one click before sending. You see the exact message, the exact recipient, the exact target before any external system writes a byte.
Connectors that read never silently start writing. Every external scope upgrade is a separate, explicit consent step. Every executed action is undoable for five minutes. Every approval and execution lands in a tamper-evident audit log. AI agency without AI roulette.
What Pulse will never do.
No individual performance scoring. No productivity rankings. No speaking-time leaderboards. No auto-scheduled meetings. No replace-the-employee suggestions. No code understanding. These are either surveillance, or someone else’s product. We will not build them.
What we will build, relentlessly, is a calm pulse of the company that admits when it’s unsure, surfaces conflicts instead of silently picking, mirrors permissions instead of expanding them, and respects the right to be forgotten by default.
Apoorv Jain · Co-founder · April 2026