For Odoo consultants and implementers

Advantages for the Odoo consultant

Not an AI module you install and forget. A business line: adapt the client’s terminology and operations — with contexts and skills. Open source (Apache-2.0); the recurring value is your knowledge.

Browse the catalog
The key moment

Generic AIs speak well… but they don’t know your client’s business rules. That isn’t in the training: the consultant brings it — with contexts and skills — and the client pays for it year after year.

Don’t say

“You need contexts.”

Say

“When the client asks for AI (and they will), either you bring knowledge of how they operate or someone plugs in a generic chat that replaces you.”

Business defense + a new service line. Contexts and skills are the customization layer of Odoo’s next cycle — like custom modules ten years ago.

Why this is good for the consultant

Six points. Detail is one click away — no idea dropped.

01

A revenue line that fits your current model

  • Project work: design and roll-out of contexts/skills per vertical or client.
  • Recurring: maintenance, expansion and tuning after every close, payroll run or season.
  • Reusable vertical packs (“ES payroll pack”, “B2B CRM pack”…) with high margin: little hardware, lots of know-how.
  • You already deliver consulting + custom. Contexts/skills are knowledge custom — not another SaaS to run.
  • Tangible deliverable: “with these 8 contexts and 5 skills, the AI does X the way your controller would”.
  • A well-built skill replicates; a prompt in an external SaaS does not.
02

Differentiation from the generic copilot — and market pressure

  • The proxy SaaS does basic ORM. You deliver “this is how we work in this company”.
  • The client stops comparing token price and starts comparing business accuracy.
  • You move from “the firm that installed Odoo” to “the firm that teaches AI to operate Odoo”.
  • Clients will ask for “ChatGPT for Odoo”. Without a serious answer, they buy the proxy.
  • Your answer: “AI yes — with your rules, on your database, with our implementation”.
  • The antidote to “a SaaS salesperson sets it up in a 15-minute demo”.
03

Your client stays with you, and the product grows with them

  • Contexts and skills are your work inside the instance: every client that uses them is a recurring service line.
  • The more institutional knowledge you embed, the less temptation to “try the cheap chat”.
  • More renewals and more work after launch day.
  • Day 1: 3 pains + a few contexts.
  • Month 6: a library of day-to-day skills.
  • Year 2: the ERP’s “brain” is a client asset… maintained by you.
04

Lower friction and risk — governance for IT and leadership

  • Data and credentials do not leave to a proxy holding the client’s passwords.
  • Writes with confirmation (governance): you can offer AI without fear it will delete something.
  • Contexts you can enable per profile or area (sales, HR, finance…).
  • IT: execution in a controlled environment, without exfiltrating data.
  • Leadership: it shows in hours and errors (less time on reports, fewer errors, consistent answers).
  • Compliance: traceability of pending/confirmed operations vs “the AI wrote via RPC and done”.
05

Faster delivery and higher consulting value

  • Fewer custom screens/reports “just to ask the same question every month”.
  • Recurring questions (rankings, mismatches, top customers…) → a skill, not an endless ticket.
  • New-user onboarding: the AI already knows the client’s jargon and rules.
  • The process workshop no longer ends as a PDF: it ends as versioned contexts.
  • The functional analyst gets a new artifact: executable knowledge.
  • Every process improvement becomes an agent improvement — visible to the client.
06

Sectors where you really win, plus a working method

  • Payroll, analytics, CRM with a definition of a “good customer”… are domain knowledge.
  • A consultant who specialises in one sector packages and bills higher than a generalist with RPC chat.
  • Barrier to entry: generic SaaS will not know your vertical; you will.
  • A method: audit a process → context → skill → test.
  • Scales juniors: seniors leave knowledge in contexts; juniors operate on top.
  • Less dependency on “the wizard who knows every domain by heart”.

Practical conversation order

1
Threat

The client will ask for AI; the proxy will own the narrative.

2
Defense

Native, governed, no passwords leaving the instance.

3
Opportunity

A new line: project + recurring + packs.

4
Proof

1 vertical / 3 pains / 2 skills in a pilot.

5
Vision

From “we install Odoo” to “we operate Odoo’s knowledge”.

Also if you build modules

You solve the domain. The AI part is already done.

You don’t build providers, history, authorization or agent plumbing. PAAP and the engine already provide that: you own your domain.

  • Plumbing solved. Providers, APIs, history, sessions and write authorization are already in the stack. You don’t reinvent them.
  • Only your domain. Shareable contexts and your own skills when needed — the rest of the agent is not your problem.
  • Apache-compatible. The reference stack is Apache-2.0. Your module may use another license on your code — see Licenses.

Inventory for module authors in Capabilities →

Prove it in a pilot: 1 vertical, 3 pains, 2 skills.

Design the knowledge, demonstrate it in the instance, and turn consulting into a recurring asset.

Trying it with a client? Tell us what you need on GitHub Discussions →

Install it →