Keep what your
experts know.
Lore interviews employees before they leave and turns their unwritten rules into context your AI agents can use.
Your agents read the wiki. The real rules were never written down.
Exceptions, workarounds, which client gets a phone call. That knowledge lives in people, and it leaves when they do.
From conversation to context.
Built to fit inside a notice period.
- 1
Interview
Lore talks with each person in plain conversation, and keeps asking until the exceptions come out.
- 2
Structure
Every answer becomes a rule with a condition, an action, a reason and the person it came from.
- 3
Connect
Your agents query Lore directly, or you export everything into the stack you already run.
Every answer shows where it came from.
Agents get the rule and the person behind it, so your team can check it, correct it and trust it.
- One endpoint, plain JSON
- A source on every rule
- Conflicts flagged, not guessed
{
"question": "Acme disputed an invoice. What now?"
}
200 OK
{
"rule": "Issue the credit, then close the ticket.",
"reason": "Acme's AP system locks closed tickets.",
"source": "Maya Chen, interview 2",
"confidence": 0.94
}The people who knew the work should be paid to teach it.
Next, we're building a way for former employees to keep teaching the agents that took over their work, on their terms and at their rate.
Someone on your team leaves soon.
Capture what they know before their last day.