Capture what's in people's heads before they leave.
11,000 Americans reach retirement age every day. When they leave, decades of institutional knowledge walk out the door.
The problem
Critical business processes exist only in the minds of long-tenured employees. Policy decisions are made in meetings with no record of the rationale. When someone leaves or retires, the team spends months reconstructing what they knew.
How it works
Here's how you build your knowledge graph step by step using the ai CLI.
Add key business processes
Start with the processes that would hurt most if the person who runs them left tomorrow.
ai add "Quarterly vendor review: 1. Pull spend data from \
NetSuite by category. 2. Flag vendors >10% over contract. \
3. Schedule reviews with top-10 by spend. \
4. Renegotiate or issue RFP for contracts expiring \
within 90 days. 5. Update vendor scorecard in shared drive. \
6. Present summary to CFO by day 15 of new quarter." \
--title "Quarterly Vendor Review" -t process -d operationsAdd policy documents
Capture not just what the policy is, but why it exists — the rationale is the part that gets lost.
ai add "Travel policy: flights over $500 need manager approval, \
over $2000 need VP approval. Hotel cap: $250/night \
domestic, $350/night international. Exceptions: \
conferences with pre-approved budgets skip individual \
approval. Policy exists because we had $400k in \
unapproved travel in Q2 2024." \
--title "Travel & Expense Policy" -t policy -d operationsAdd decision rationale
Decisions without documented rationale get relitigated every 6 months. Capture the why.
ai add "Decided to keep EU data in Frankfurt region only \
(not multi-region) because: 1. GDPR data residency \
requirements from our top 3 EU clients. 2. Latency \
acceptable (<100ms) from single region. 3. Multi-region \
would add $180k/year in infrastructure costs. \
Revisit if we sign APAC clients." \
--title "EU Data Residency Decision" -t decision -d strategyAdd vendor and partner context
The institutional knowledge about vendor relationships, contracts, and contacts is some of the first to disappear.
ai add "Stripe relationship: Enterprise plan, negotiated \
2.4% + 30c (standard is 2.9%). Contact: Jane Smith, \
Account Exec. Contract renews Nov 2026. \
Key integration: webhook signing uses v2 schema. \
Escalation path: jane@stripe.com -> their VP of Sales." \
--title "Stripe Vendor Relationship" -t note -d operationsAdd compliance requirements
Compliance knowledge is specialized and expensive to reconstruct. Capture it proactively.
ai add "SOC 2 Type II: annual audit in Q1, auditor is Deloitte. \
Key controls: access reviews quarterly, MFA enforced, \
change management via PRs with approval. \
Evidence collection starts 30 days before audit. \
Owner: Security team. Budget: $85k/year." \
--title "SOC 2 Compliance Process" -t process -d operationsLink processes to policies
Connect everything so that when leadership queries the graph, they get the full picture.
ai link abc123 def456 --rel "governed-by"
ai link ghi789 jkl012 --rel "requires""Vendor Review" is governed by "Travel & Expense Policy". "SOC 2 Process" requires "EU Data Residency Decision".
Leadership queries before decisions
Before making a decision, query the graph for prior decisions, policies, and context.
ai context "What do I need to know about our EU data strategy?"Returns the data residency decision, compliance requirements, relevant vendor contracts, and their connections.
Institutional knowledge is preserved
Business processes, policy rationale, and vendor relationships are captured in a connected knowledge graph. When people move on, the knowledge stays.
Critical processes documented before the people who run them leave
Decision rationale prevents re-litigation of settled questions
Compliance knowledge is searchable and connected to processes
Leadership gets full context before making new decisions
Start building your knowledge graph
Free during beta. No credit card required.