Engineering and technical consultancies

AI for engineering firms: 3 published cases, and what stays with the engineer

In an engineering firm, AI helps prepare quotes, qualify prospects and find information in standards and documents, while the engineer keeps the costing and the signature. Codito Ergo Sum, a French AI agency, has published 3 cases in these trades: GAEA Géotechnique, AXS Ingénierie and Jarvis. It works through the AI Partnership from €499 excl. VAT per month, no commitment, and through development on quotation, from €3,000 excl. VAT.

Book 30 minutes

3 published cases in engineering firms

Our 3 published cases concern quoting, prospecting and working with technical documents. In each, a person from the trade has the final word.

  • GAEA Géotechnique, a geotechnical consultancy founded in 2019 in Peypin, producing the reports for assignments G1 to G5 of the French standard NF P 94-500. In the AI Partnership since May 2026. Its agent prepares G2 study quotes from its Excel catalogues and submits them to the firm. It was moved to Claude Code in September 2026.
  • AXS Ingénierie, an engineering firm in Le Havre specialising in heavy lifting equipment and industrial and port installations. Its prospecting agent is in service: it ranks companies by their resemblance to clients already won, explaining each score, using a sales history updated every night. The quote-drafting agent is the next step.
  • Jarvis, for landscaping and utilities studies: 4 specialist agents in production (landscaping, utilities and roadworks, hydraulics, technical specifications), each with its own knowledge base, handling plans and diagrams and exporting to PDF or DOC.

Where AI helps along a project

AI helps at every stage where the engineer searches for, rereads or reformats information that already exists. At GAEA, 3 tasks consumed the team's time: analysing documents received, writing reports and searching for information.

  • Before the project: identifying the right prospects, spotting buying signals, suggesting an approach.
  • At quoting: structuring a quote from your catalogues, leaving the price to the engineer.
  • At analysis: reading specifications or a file received and flagging points to check.
  • During the study: finding a standard, a chart or a regulatory passage, with its source.
  • At drafting: preparing the repetitive parts of a report in your templates.

What stays with the engineer

The engineer keeps the calculation, the costing, the interpretation and the signature, since they carry the firm's liability. Our clients set these limits from the start.

  • At AXS, costing remains with the engineers.
  • At GAEA, nothing is sent automatically: human validation remains systematic.
  • A cited standard is checked in the reference text, as a model can be wrong while giving a reliable-looking answer.
  • Data on your clients and projects falls under your contracts: choose the model and hosting accordingly. AXS had its agents delivered on its own server.

What we cannot tell you

We cannot give you a quantified gain on these 3 cases, as none has been published to date. We prefer to write this here than to put forward an unfounded estimate. To measure what it would change for you, the calculation is done on your own tasks, with your teams, using the formula: time lost × loaded hourly rate × annual recurrence.

Which format, and at what price

The format depends on who will take over the agent: your team in the AI Partnership, or our team in development. All prices exclude VAT.

  • Training the team to build its agents, as GAEA set out to do: Premium AI Partnership, €799 per month, no commitment, 4 hours per month split as you wish.
  • Having a specific agent built from your specifications, like AXS: Custom Development, on quotation, from €3,000, with 1 to 2 weeks of scoping.
  • Quantifying your frictions before deciding: AI Audit, Express at €2,000, Focus at €5,000, Essential at €7,500, Max at €10,000.

Who this is not suited to

This is not suited to you if you expect an agent to carry out or sign calculations for you, or if your catalogues and templates do not yet exist in usable form: they must be created first. Nor is it suited to you if your volume of quotes remains too low to make a tool pay.

If you are a construction contractor and not an engineering firm, Codito BTP was designed for you. In other cases, book 30 minutes, free of charge, with an example of a quote or report in mind.

Your questions

How can AI be used in an engineering firm?

Start with a repetitive task based on your documents, such as preparing quotes from catalogues: this was GAEA Géotechnique's first project. The team validates each result, and no quote goes out automatically.

Can AI replace the engineer on costing?

No, in our published cases, costing and the decision remain with people from the trade. At AXS Ingénierie, the planned drafting agent will write the technical quote, but costing remains with the engineers.

What quantified results do you have for engineering firms?

We have not published any quantified gain on these cases. The only published volume is that of Jarvis: 4 specialist agents in production. The gain is calculated on your own tasks, with your teams.

How much does an AI agent cost an engineering firm?

Your team can build a simple agent in the Premium AI Partnership, at €799 excl. VAT per month, no commitment. An agent developed by Codito is quoted from €3,000 excl. VAT, after 1 to 2 weeks of scoping.