INTELLIGENCE LAYER

Your AI tools forget everything. Indigo doesn't.

Every AI session starts from zero. Indigo captures signals from 6 channels — meetings, email, Slack, GitHub, Linear, docs — structures them into decisions, actions, and insights, and makes them available to any agent through your filesystem, MCP, or API. The more you use it, the smarter every agent gets.

SIGNAL CHANNELS

THE INTELLIGENCE LAYER

Signals. Context. Action.

Two paths converge into one intelligence layer any agent can read. Hover any node to explore the data flowing through it.

S — SIGNAL STREAM (HOT)

Meetings

Live transcripts from Zoom, Teams, Google Meet

Speaker diarization

Action item detection

Decision extraction

Chat & Email

Slack threads, email chains, and direct messages

Thread context preserved

Attachment indexing

Sentiment signals

Project Trackers

GitHub PRs, Linear issues, Jira tickets

Status change tracking

Cross-reference linking

Commit attribution

live

Insights Agent

Extracts, classifies, and routes signals in real-time

0.94 precision on decision detection

< 2s from event to structured signal

Decisions · Actions · Accomplishments

DATA OBJECT

{
  "type": "signal",
  "source": "slack/#eng-standup",
  "extracted": {
    "decision": "Use Redis for session cache",
    "owner": "Maya Chen",
    "confidence": 0.94
  }
}

structured

Decision Ledger

Structured record of every decision with full provenance

~2s recall latency

Full source attribution chain

Time-stamped audit trail

ISOLATED DATA LAYER

Isolated Cloud Bucket

Per-company encrypted storage — cloud-hosted, locally synced

Company-scoped encryption (BYOK on Enterprise)

Local sync daemon for offline access

SOC2 Type II isolation boundaries

Data residency controls

C — CONTEXT STREAM (COLD)

Wikis & Docs

Notion, Confluence, Google Docs, internal wikis

Version-aware indexing

Structure preservation

Auto-refresh on edits

Policies & Specs

SOPs, compliance docs, technical specifications

Regulatory documents

Technical specs

Process documentation

Strategy Memos

Board decks, strategy docs, quarterly reviews

Board narratives

OKR documents

Competitive analysis

nightly

goClaw · Context Engine

Autonomous research agent — crawls, synthesizes, builds knowledge nightly

Autonomous domain crawl

Nightly synthesis → knowledge cards

Context compounds over time

DATA OBJECT

{
  "type": "knowledge_card",
  "domain": "competitor/acme-analytics",
  "title": "Acme Q3 pricing change",
  "summary": "Raised enterprise tier from $2k to $3.5k/mo...",
  "confidence": 0.88
}

synthesized

Evidence Store

Persistent, evolving institutional memory

Citation-backed entries

Self-evolving corpus

Full provenance chain

A — ARCHITECTURE (FUSION)

Deep Research Agent

Model-agnostic agent querying both streams simultaneously

Hot + cold path fusion

Claude, GPT, Gemini, open-source

Answers · Drafts · Actions

DATA OBJECT

{
  "type": "query_result",
  "query": "What did we promise Acme Corp on BYOK?",
  "sources": {
    "signals": 847,
    "knowledge_cards": 23
  },
  "answer": "Three commitments found across email, call, and Linear...",
  "citations": 3
}

Read the SCA paper →Explore the Context Engine →

ACCESS MODES

Works with the agents you already use

Access the intelligence layer via CLI, REST API, or MCP server. Agent-agnostic, model-agnostic — same data, your preferred interface.

CLI

Query signals directly from your terminal. Pipe results into scripts, dashboards, or any downstream tooling.

indigo signals --channel meetings --since 7d

REST API

Programmatic access to your full signals corpus. Filter by channel, date range, or custom tags.

curl https://api.getindigo.ai/v1/signals \ -H "Authorization: Bearer $INDIGO_KEY" \ -d '{"channels": ["meetings", "slack"]}'

MCP Server

Expose your signals to any MCP-compatible AI client. Claude, Cursor, and others can query your data natively.

// claude_desktop_config.json { "mcpServers": { "indigo": { "command": "indigo", "args": ["mcp", "--port", "3001"] } } }

AGENT AMNESIA

Every AI session starts from zero. Indigo gives your agents memory.

What did we promise Acme Corp on BYOK — pull signals from email, calls, and Linear since Aug 1.

What did Customer Success signal about BetaCorp churn risk — across email and Slack — in the last 5 days?

What signals came in from Engineering this morning across Slack, GitHub, and Linear?

Which meeting signal contradicts what was committed in the GitHub PR description last week?

Which Linear issue has conflicting signals — one owner says done, another says blocked?

What signal from the Sep 12 email thread kicked off the pricing change — link to original source.

PRICING

Pay for what you process.

Start free. Scale as your signal volume grows. No seats, no surprises.

Free

500 signals / mo

no credit card required

Signal ingestion (up to 500/mo)
Decision Ledger access
CLI + API access
7-day data retention

Pro

$0.002 / signal

volume discounts at scale

100K signals = $200 · 500K = $850 · 1M+ = contact us

Everything in Free
All 6 signal channels
Unlimited data retention
Team collaboration
Priority support

Enterprise

Custom

Everything in Pro
BYOK (data + models)
Custom integrations
SLAs + uptime guarantees
Dedicated deployment
Compliance + audit logs

Signals include meetings, emails, GitHub events, Linear updates, Slack messages, and custom webhooks.

Start capturing signals

Ready to give your AI agents memory? Free tier, no credit card required. Your Decision Ledger is live in minutes.