Backed by Relevance Lab· previously known as Kaman.aiInsightsWhitepapersTrust centreCareersContactSign in
Try it now

Enterprise AI platform, products and services

AI agents
you can let act.

Camaan puts AI agents to work on your data, under rules you set for every kind of action, with proof of every step. Delivered with engineers from our services partner, Relevance Lab, and yours to own: source available, under your brand, on your premises.

Talk to our teamSee how the platform works
Blog

Proof, not logs: what an auditor needs to see when an AI agent acts

A log says something happened. Proof lets someone who wasn't there check what happened, why it was allowed, who approved it, and that nobody changed the story.

9 min read Point of view

The forward-deployed trap: why most agents built for you get abandoned

8 min read New

The Camaan Marketplace: 23 offers by process, team, platform and industry

Browse the offers
Per actionAutonomy you set for each kind of action: alone, ask first, or never.
Every stepSigned with its reason, its rule and who approved it, in a ledger you can show a regulator.
Same answerScenarios fix the data, steps and rules, so recurring work repeats.
Your boundaryOur cloud, your cloud or your servers, with the models you choose.

Organisations putting agents to work with Camaan and Relevance Lab

Read their stories
  • ABeam Consulting
  • BigTapp
  • Apar
  • NexusNow
  • DropSkip
  • Traxon e:cue
  • Action Lab 2050
  • Reality Scale
  • AgileIntel Research
  • zclap
  • The Maze
  • Oceanova Exports

Our perspective

The question is no longer whether AI can do the work. It is whether you can let it, and prove what it did.
Gartner predicts over 40% of agentic AI projects will be cancelled by the end of 2027, citing cost, unclear value and weak risk controls (June 2025).

Autonomy is a policy, not a prompt

Decide, per kind of action, what an agent may do alone. Reading a record is not the same as moving money, and the rules should say so.

Repeatable beats impressive

Regulated work needs the same answer to the same question. Fix the steps, data and rules once; let the model fill in only what changes.

Proof, not logs

An audit trail you can't verify is a story. Every action and decision should be signed, chained and checkable later.

Own the asset

Your data, your ontology, your agents and your code should stay yours, under your brand, wherever you choose to run them.

Why now

The market, in four numbers.

Enterprises are moving from pilots to programmes, and most of what was built so far won't make the move. What survives is packaged, governed and owned.

95%

of generative AI pilots showed no measurable profit-and-loss impact.

Packaged offers with proof, not open-ended pilots.MIT NANDA, The GenAI Divide, Aug 2025 · source
40%+

of agentic AI projects will be cancelled by the end of 2027, for cost, unclear value or weak risk controls.

Rules per action and a signed record, from day one.Gartner, June 2025 · source
70%

of enterprises will abandon agentic AI built for them by vendors' forward-deployed engineers by 2028. Under 20% of those engagements turn recurring needs into product.

Offers built once, and a layer you own.Gartner, September 2026 · source
$37B

spent on generative AI by enterprises in 2025, about three times 2024.

Buyers want offers they can buy, not projects.Menlo Ventures, Dec 2025 · source

Customer stories

Agents at work, on real operations.

From supply-chain planning to fire watch in a forest reserve. All customer stories

Supply chain · Redefined Logistics, New York5 systemsstore, ERP, warehouse, carrier and broker, read into one planning model

A supply-chain decision platform, built entirely on Camaan.

Read the storyPublic sector · A national wildlife sanctuary, India12 + 75AI cameras and trap cameras under contract, every alert reviewed by a ranger

Fire watch across a forest reserve, from camera towers to the control room.

Read the storyNon-profit · Action Lab 20507 tablesbuilt from census tables, survey reports and PDFs, every figure traced to source

From public reports and private files to tables a policy team can defend.

Read the story

Camaan in numbers

Counted, not claimed.

Counted from the platform itself in November 2026.

105connectors to the systems you already run, from SAP and ServiceNow to IBM i
23ready-made offers in the marketplace, each tagged by process, team, platform and industry
12writing systems read by the built-in OCR, on your own servers
373user journeys proven end to end in a real browser before release
1,875proof screenshots kept with those journeys

What we do

One platform, the products built on it, and the people to deliver it.

Most enterprises buy a tool, then hire someone to make it work. Camaan and Relevance Lab bring both, and leave you owning the result.

Platform

Build, run and govern AI agents, workflows and repeatable scenarios on your own data.

  • Agents, workflows, scenarios and skills
  • Lake engine, ontology and knowledge graph
  • Per-action autonomy and the audit ledger
  • Bring your own models and engines
Explore the platform

Products

Ready-made products for specific jobs, built on the platform and governed the same way.

  • Camaan Twins
  • Camaan Fletcher
  • Camaan Voice
  • Camaan Apps
See the products

Services

Relevance Lab's engineers work alongside your teams, from the first use case to a platform your people run.

  • Assess where agents will pay off
  • Prove it in a pilot on your data
  • Build and scale across teams
  • Run it for you, or hand it over
How we work with you

The platform

Everything an agent needs, in one governed stack.

Each layer is part of Camaan rather than another product to integrate. Governance runs through all of them, and Relevance Lab's services wrap the whole.

The Camaan stack: products on top of channels, build, knowledge and data, runtimes and deployment, with governance across every layer

Proof, not logs

Every action says how it was decided. Right where it happened.

Under each reply, a line per action: allowed by which rule, approved by whom and when, or refused and why. One click opens the signed record, and Verify checks it.

See what the record holds

camaan.ai / chat
✓List files was allowed under "read actions run without asking for this agent"
✓Render scenario was approved by Priya Nair at 14:02 under "execute actions ask before running for this agent"
✕Delete file was denied by "destructive actions are forbidden for this agent"

Products built on Camaan

Ready for specific jobs. Governed like everything else.

Each product shares your data, your rules and your audit ledger, and runs wherever the platform runs.

Built on Camaan

Camaan Twins

Live models of plants, fleets and networks, fed by their sensors and systems. Ask what if, and try a change on the copy before the real thing.

For operations, engineering and asset teams

Explore Camaan Twins
Built on Camaan

Camaan Fletcher

Journey-driven software delivery. Agents build each change on your machines, a real browser test proves it, and a person decides what merges.

For engineering leaders and delivery teams

Explore Camaan Fletcher
Built on Camaan

Camaan Voice

Agents that answer and place calls on your existing phone lines, in many languages, under the same rules and audit as every other channel.

For contact centres and service teams

Explore Camaan Voice
Built on Camaan

Camaan Apps

Describe an internal app and build it with AI: versioned in git, previewed safely, tested end to end, and deployed on the platform.

For business and IT teams

Explore Camaan Apps

Marketplace

One engine. Knowledge packaged on top.

23 ready-made offers, each described by the process it runs, the team it serves, the platform it runs on and the industry it knows. Install one, and it is yours to change. Browse the marketplace

Find offers for issue to resolution that serve operations on ServiceNow in any industry

Own it

Your platform. Your brand. Your premises.

Use Camaan as a service, or take it in-house as your own AI platform: the code, the brand and the infrastructure all under your control.

Source available

Read, audit and extend the code under licence. Your security team can see exactly what runs.

White-labelled as yours

Your name, logo, colours and domain. Your people see your company's AI platform.

On your premises

Runs on your own servers with the models you choose, so your data stays inside your boundary.

Three ways to deploy: Camaan cloud, your cloud account, or your own premises under your brand
Camaan provides the platform, Relevance Lab the engineers, and you the data, rules and brand; together, agents you can let act

Camaan and Relevance Lab

A platform company and a services company, on one team.

Camaan builds the platform and the products on it. Relevance Lab, our services partner and investor, brings the engineers who deliver it with you: assessing where agents pay off, proving it on your data, and handing over a platform your people run.

  • One plan for the platform and the delivery
  • Delivery engineers trained on the platform, working with ours
  • A plan to hand it over, from the first day

About Camaan and Relevance Lab

Industries

Where accountable AI matters most.

Typical places we start. Every one involves actions someone must be able to answer for. All industries

Financial services

  • Onboarding and review workflows
  • Payment and refund approvals
  • Reporting from governed data

Life sciences

  • Document review and summarisation
  • Supply network visibility
  • Repeatable analysis

Energy and utilities

  • Twins of plants and networks
  • Field and maintenance workflows
  • What-if before a change

Insights

What we're learning from putting agents to work.

All insights

Blog

One engine, knowledge on top: how we package agents for the enterprise

Agent platforms are converging. What a buyer can't copy is knowledge of an industry, a process, a team and a platform, built once and delivered at the depth they need.

10 November 2026 · 10 min readBlog

Finding the revenue that leaks between order and cash

5 November 2026 · 8 min readBlog

Teaching agents your business: building an ontology by talking it through

3 November 2026 · 7 min readPoint of view

Agents you can let act: a governance model for autonomous AI

20 October 2026 · 12 min read

Trust

Security your auditors can check.

Visit the trust centre

ISO/IEC 27001Our information security management is certified.
Signed audit ledgerEvery action and decision, chained and verifiable.
Access in the enginePeople and agents query as themselves; row rules and masks apply in the engine.
Secrets in a vaultCredentials are never stored in tables or settings.
Your boundaryRun on your servers with your models when data cannot leave.

Tell us about the work you'd like an agent to do.

Talk to our team