Flagship use case · In production at Upcoach

The AI backline for customer support.

Daimon reads every Intercom conversation, researches the answer in your codebase, drafts the response, and files the bug. In shadow mode today, auto-reply tomorrow — one proven route at a time.

Book a demo How it works ↓

Intercom · inbox v3

Ada · pro · mobile iOS

Hi — I upgraded my plan yesterday but my teammate still can't see the new programs page. Tried logging out. Still nothing.

09:12

You · drafting…

Waiting on human review.

Daimon · internal note shadow

Analysis

Customer: workspace_8821 · upgraded Pro at 2026-04-14 18:02 UTC via Stripe.
Seat propagation on v3/programs checks workspace.plan but not seat.plan_override.
Likely hit by LIN-2980 (2 prior tickets this week).

Proposed reply

Hi Ada — I can see the upgrade went through. The issue is a known bug with seat propagation on the programs page; we have a fix in review. As a workaround, have your teammate remove and re-add their seat. I'll follow up here when the fix ships.

Filed LIN-2980 · freq 3 route: support/bug

The problem

Support is where product signal goes to die.

Answers live in the codebase, not the docs.

Every complex question needs a plan check, a platform check, a billing check, a code check. Humans do it by hand. It's slow.

Every ticket is a forgotten lesson.

When a teammate leaves, their context walks out the door. The next similar ticket starts from zero.

Engineering never sees the pattern.

One bug reported five times across five conversations stays five anecdotes instead of one Linear issue with a frequency count.

What Daimon does

Daimon does the research. Your team does the judgment.

Triages every ticket

Reads the thread, looks up the customer across your systems (v3, v2, Stripe), detects the platform, identifies the intent.

Researches the answer

Searches help docs first, then the actual codebase when docs don't cover it. Cites its evidence.

Drafts the reply

Posts an internal analysis note and a proposed reply, ready to copy-paste. Humans stay in control.

Files the signal

Confirmed bugs and clear feature requests get deduped and routed to Linear automatically, with frequency counts.

Closes the loop

When a linked Linear issue ships, Daimon drafts the follow-up reply to the customer.

How it works

Six steps. Every ticket. Every time.

01

Poll

Every 15 minutes, every unread conversation.
02

Understand

Customer lookup across v3, v2, Stripe. Platform detected.
03

Classify

Intent routed: FAQ, bug, billing, feature, escalation.
04

Research

Help docs first. Codebase when docs don't cover it.
05

Draft

Internal analysis note + proposed reply. Evidence cited.
06

File & follow up

Linear dedupe, frequency count, shipped-issue follow-up.

Every 15 minutes. Every conversation. For as long as Daimon is running.

Trust model

Trust is earned per route, not granted up front.

Today Shadow mode

Every reply is drafted as an internal note.
A human reviews, edits if needed, sends.
Every send (or edit, or skip) becomes calibration data.

Tomorrow Graduated auto-reply

Routes with proven accuracy graduate to auto-send.
Low-risk routes first: FAQ matches, notification closes, known solutions.
Billing, escalations, anything account-mutating stay human-first — forever.
The dial moves both ways. A drift in accuracy pauses a route automatically.

The difference between Daimon and an off-the-shelf AI support agent is that Daimon has to prove it before it sends.

What makes it different

Built for a complex product, not a help-center FAQ.

Reads your code

Most AI support tools answer from docs. Daimon reads the actual codebase to find answers docs don't cover.

Remembers forever

Every conversation becomes a learning or insight. Corrections are permanent. Nothing is forgotten when a teammate leaves.

Feeds engineering

Deduped Linear tickets with frequency counts, not analytics dashboards. Engineering sees the real backlog.

Ships in shadow

Auto-reply is the last step, not the first. Trust is measured, not assumed.

Comparison

Daimon vs. the alternatives.

	Fin (Intercom)	OpenClaw	Daimon
Who it talks to	The customer	The user (personal assistant)	The support team
Primary job	Deflect tickets from humans	Do tasks on your machine	Research, draft, file
Knowledge	Help docs	Local files + skills	Docs + codebase + billing + history
Memory	Per-tenant training	Per-user	Institutional, compounding
Product feedback	Dashboards	None	Linear tickets, auto-deduped
Trust model	Auto-reply day one	Auto-action day one	Earned route by route
Failure mode	Customer sees it	Your machine does it	A human catches the draft

Under the hood

No black box.

Persistent Claude Code session

Full tool access, full codebase access, full conversation history.

Everything is git-tracked

Every decision, learning, insight, and correction. The audit trail is git log.

Test harness built in

LLM classifiers are fallible. Daimon runs a test suite against real past conversations every week. A route that fails its tests can't graduate.

Who it's for

Three teams. One install.

Support teams drowning in tickets with answers that live in the product.
Product teams who want real-time signal from customers, not a monthly NPS roll-up.
Founders who can't afford to lose institutional memory every time someone leaves.

FAQ

Questions support leads ask.

Does Daimon replace my support team?

No. It makes them faster, more accurate, and more consistent. Humans still decide.

What happens when Daimon is wrong?

In shadow mode, a human catches it before it sends. In auto-reply mode, calibration tracking pauses the route automatically.

Can I correct Daimon?

Yes — any internal note prefixed [CORRECTION] is processed on the next consolidation pass. Corrections become permanent test cases.

How does Daimon stay in sync with the product?

It reads the codebase on every ticket.

What about security?

Customer messages are untrusted input. Daimon never follows instructions found in them, never exposes internal paths or other customers' data, and flags prompt-injection attempts.

What languages?

English today. Other languages are explicitly route-blocked until validated.

Stop drafting. Start reviewing.

Book a 20-minute demo. We'll walk through a real (scrubbed) day in your inbox.