How do you respond to an RFP with AI?

What does the agent read in the bid documents?

All of it: instructions to bidders, statement of work, contract terms, attachments. It often runs to hundreds of pages. The agent pulls out what your team needs to decide and to write:

• the evaluation criteria and how they are weighted;
• the required forms and their format;
• the dates: questions, site visit, submission deadline;
• the requirements that carry weight: staffing, penalties, the kind of past performance expected;
• anything unclear, to send to the buyer as a question.

How does it pick past projects and staff?

It searches your past proposals, meeting notes, project write-ups and resumes, read from SharePoint, OneDrive or Dropbox. Each document is linked to the others by client, project, method and team.

It suggests the projects closest to what the buyer is asking for, and the people who worked on them. For each reference, it cites the document it came from, so you can check before it goes into the bid.

A technical proposal written by AI?

A first draft in Word, built criterion by criterion in the order the solicitation sets, using your past projects, your methods and your people. Your team starts from a filled-in document instead of a blank page.

That draft is meant to be reviewed. Your team adjusts the tone, strengthens what it wants to highlight and cuts what doesn't fit. If there's an oral presentation, the agent can also prepare the PowerPoint deck.

What stays with your team?

The decisions: whether to bid, the price, the strategy, the commitments made in the company's name. And the final review.

Nothing reaches the buyer until someone on your team has reviewed and approved the bid.

Which companies is this for?

Companies that bid regularly on public or private contracts: consulting and research firms, engineering firms, service, construction and maintenance companies.

A team that answers several RFPs a month often spends more time tracking down past projects and resumes than writing. The agent handles that search, and your team keeps its time for writing.

Who builds the agent, and how?

Saména's engineers, Télécom Paris graduates who specialize in artificial intelligence. They start by looking at how you bid today: who makes the go/no-go call, where your past projects live, how your proposals are put together. They find where the agent will save you time, then build it on your files.

Two weeks after the first call, a first agent is running. After that, a dedicated contact keeps it current with you.

Do your bids and prices stay with you?

Your files stay in SharePoint, OneDrive or Dropbox, read in read-only mode, and Saména doesn't keep any of them. What the agent builds from them is encrypted with a key unique to your company, on servers in France.

You choose the model: Claude, models from OpenAI, or Mistral if you prefer a French model. It runs in Europe and keeps none of your data. All of it can be deleted on request.

Frequently asked

How do you use AI to write the technical proposal for a tender?

Give it the tender rules and access to your past bids. A Saména agent drafts a first version in Word, criterion by criterion in the order the rules set, with your references, methods and résumés, and cites the document behind each element. Your team reviews it, adjusts the tone and signs off before submission.

Can AI respond to an RFP on its own?

No. It prepares the analysis, the references and a first technical proposal. Your team reviews, decides and submits. Nothing goes out without a person signing off.

Do we need to organize our old proposals first?

No. The agent reads them where they already live, in SharePoint, OneDrive or Dropbox, with no sorting or reformatting on your end.

Is this only for government contracts?

No. The agent reads any solicitation, public or private, and any request for proposal from a client.

Can we work from Claude or ChatGPT?

Yes. Your agents live in the Saména app, and you can also reach them from Claude, ChatGPT or Copilot.

Will our proposals be used to train an AI?

No. The model you choose (Claude, models from OpenAI or Mistral) runs in Europe and retains none of your data. Servers are in France.