Toucan · Custom analytics, standalone or embedded

Custom analytics, delivered in the AI era.
Standalone or embedded in your product.

Dashboards and data apps your teams and your customers actually use.

or generate a data app yourself, free

A live data app built from your website in ~2 minutes.

Live example · a custom data app
green-got.toucan.cloudopen ↗
Trusted by teams at
Willaman ReachFive Georgetown University Crédit Agricole Combo BNP Paribas

Two offers, one platform

For business leaders Custom Apps · fully managed
A custom data app, built for you.

You want a data app for your business, without building a team. Our Crew builds it and runs it for you, and keeps improving it.

  • Fully bespoke, shaped for how your business decides
  • Supervised AI agents build it, no work on your side
  • A human Toucan Lead signs every release
  • You follow and request changes from the platform
For product & data teams Platform · self-service
A platform to build your own.

You want to build dashboards and embed them in your product. The platform gives you the tools, you keep the control.

  • A serious semantic layer: metrics, joins, calculation rules
  • AI that answers questions and builds dashboards on request
  • A rich visualization library, ready to embed at scale
  • Row-level security, embedding, your data sources

The problem

Analytics for the AI era.
Every other route falls short.

BI is slow and rigid

Existing BI bolts on an AI button, but the roadmap crawls and the model stays boxed in. You adapt to the tool, not the reverse.

A raw LLM is ungoverned

An LLM on your data is fast, but its answers come with no permissions, no business context, no audit trail and no one accountable.

Building in-house is slow

Hiring a data team and wiring a stack takes months before the first useful screen. The blank page is the expensive part.

What makes Toucan different

The parts that actually matter.

Whichever offer you choose, the same engine runs underneath.

Ask your data

Ask a question. Get the answer.

Type it the way you'd say it out loud. Toucan reads your data and answers in seconds, with the chart to back it up. Anyone can do it, straight from the app.

  • Plain language in, chart and numbers out
  • See the data behind every answer
  • No SQL, no waiting on the data team
AIManual
employees by age
Completed in 9 seconds
Done. I added an "Employees by Age Range" bar chart to your dashboard, showing all 60 employees across age groups.
36–45 · 18 employees
56–65 · 18 employees
46–55 · 10 employees
Employees by Age Range
6
22–25
8
26–35
18
36–45
10
46–55
18
56–65
+ Add to dashboard
What dashboard would you like to build?
Semantic layer

One definition. Everywhere.

Set your metrics once. Revenue, active client, margin, all defined in one place. Every chart, every answer, every app uses the exact same numbers. What you see is always what you meant.

  • Metrics, joins and rules in one governed place
  • The AI answers from your definitions, not guesses
  • Change it once, it updates everywhere
net_revenuegross − refunds − credits
active_clientbilled in last 90 days
gross_marginnet_revenue − cogs
churn_ratelost / active, monthly
used by 24 dashboards · the AI · 3 embedded apps
AI context & memory

It knows your business.

Teach it your rules once and it remembers. Your definitions, your exceptions, your fiscal year. Correct it once, it stays corrected. It gets sharper every time you use it.

  • Set your rules in plain language
  • Corrections stick for good
  • Feed it glossaries and business context
"Active client" = billed in the last 90 daysremembered
Exclude non-compliant accounts from revenueremembered
Our fiscal year starts in April, not Januaryremembered
"EMEA" includes the UK for sales reportingjust taught
Multi-tenant & row-level security

Embed once. Everyone stays separate.

Ship analytics inside your product, in your brand. Every customer sees their own numbers and only theirs, on dashboards and on AI answers alike. Scoped by default.

  • One deployment, every customer isolated
  • Row-level security on AI answers too
  • Your brand, your domain, their own login
Customer analyticsAcme IncEmbedded
Revenue
€2.4M
Active users
1,204
Retention
94%
Usage by month · Acme Inc only
Apr
May
Jun
Jul
Aug
Sep
🔒 Scoped to Acme Inc · 1,204 of 48,000 rows visible

Two ways to get there

For business leaders Custom Apps · fully managed
Have it built for you.

Tell us what your business needs. Our Crew builds the app, runs it, and keeps improving it, with a Toucan Lead accountable for every release.

For product & data teams Platform · self-service
Build it yourself.

Get the platform and build your own dashboards, then embed them in your product. You keep the control, the engine does the heavy lifting.

How it compares

Built for the gap legacy BI and generic AI left behind.

The honest version. Where each kind of tool is strong, and where it leaves you stuck.

ToucanManaged app or platform Legacy BI
Power BI, Tableau
Modern BI
Looker, Metabase
Generic AI
ChatGPT on your data
Ask in plain language, and trust the answer Answers from your semantic layer, with the data shown No natural language Add-on, varies by tool Fast, but guesses joins and invents metrics
Knows what your metrics mean Semantic layer, defined once Modeled by hand per report Semantic layer, if you build it No business context
Embed in your product, per-customer isolated Yes, row-level security on every viewer Costly add-on Possible, engineering-heavy Not designed for it
Time to a first useful result Days Weeks of setup and modeling Weeks, plus a warehouse Minutes, if the answer is even right
Who builds and maintains it You, or our Crew for you Your team Your team You, prompt by prompt
Governance: row-level security, SSO, audit Included Enterprise tier Enterprise tier None

Two offers. One platform.

Have a custom data app built for you, or build your own on the platform. Both embeddable, both on the same engine.