NewOcto Context Engine

Your AI now knows your company.

Context, not content. The Octo Context Engine brings your company context from the Octopus app straight into Claude or ChatGPT.

Request a demo→Take the quick test
How your AI works with the Octo Context Engine. Illustration with sample data.
Octopus app
Your company context
  • Purpose
  • Values
  • Core customers
  • KPIs
  • Role scorecards
  • Blockages
Your AIClaudeChatGPT
|
reads your contextwrites back once confirmedIllustration with sample data
The problem

An AI without context is like a brilliant consultant on their first day.

Smart, but they don't know you. Ask them for a job ad for a head of sales and you get a text that fits any company. To change that, you'd have to explain who you are from scratch with every question.

Nobody does that.

Job ad: head of salesWithout context

We are an innovative, growing company.

Exciting tasks await you in a dynamic team.

We offer an attractive salary and flat hierarchies.

Fits any company.
Job ad: head of salesWith the Octo Context Engine
Purpose
Why you exist
Role scorecards
What the role must deliver
Core customers
Who you work for
Values
How you work together
Written for exactly your company.
The solution

The Octopus app is this consultant's onboarding and the notebook they write back to.

Through an MCP server, an open standard interface, your AI reads your purpose, values, core customers, KPIs, role scorecards and your blockages. For the job ad, that means it's written for exactly your company.

Whatever the AI adds, it writes back to the app once you confirm.

How it works

  1. 1

    Connect

    You connect your AI to the Octo MCP server.

  2. 2

    Sign in

    You sign in with your Octo account and only see the companies you've been given access to.

  3. 3

    Choose a company

    You pick the company, and your AI works live with your context.

The AI suggests changes, and a human confirms them.

Where it helps

Monthly figures

Enter the September figures. Where are we against plan and last year?

usesKPIswrites back once confirmed

Check a prospect

A machine builder with 80 employees got in touch. Do they fit our core customer? If so, write a first message.

usesCore customers

Performance review

Prepare my annual review with our head of customer service, based on her role scorecard.

usesRole scorecards

Workshop notes

Check my notes on our core customer against our niche and value proposition and show me contradictions.

usesCore customers

Everyday problem

Our meetings end without a decision. Which blockage is behind it, and what should I try?

usesBlockages

After the meeting

Here's the transcript of our monthly meeting. Summarise the decisions, assign each task to the responsible role and add the figures we discussed to the KPIs.

usesRole scorecardsKPIswrites back once confirmed
And for your culture

Your AI knows your values, too.

And the behaviour examples that go with them: what's cool with you and what isn't.

Values dilemma

A top salesperson regularly breaks agreements. What do our values say about that?

Feedback

Help me phrase my feedback to the team along our values.

Sharpen values

Suggest concrete examples for our value reliability, cool and uncool.

Once you confirm, the AI adds them to the app.

  • Philipp von Kunhardt
    Philipp von Kunhardt
    Partner · AI expert
  • Dr. Patrick Blum
    Dr. Patrick Blum
    Partner · AI & learning expert

We'd love to show you the Context Engine ourselves.

Request a demo, or take the free quick test first.

Request a demo→Take the quick test