CNPq (Brazil’s National Council for Scientific and Technological Development) has opened public call 29/2026 under the RHAE programme (Recursos Humanos em Áreas Estratégicas, human resources in strategic areas), dedicated to artificial intelligence, with applications running until October 9, 2026: the grant pays a researcher to work inside a company, not inside a university.
What RHAE pays for
RHAE funds fellowships for a researcher working inside a company and attached to that company’s project. In this call, which is specific to artificial intelligence, the fellowship types are SET, DTI, EV and ATP, at every level. A project may run for up to 30 months, and each company may raise up to R$ 300,000 in technological development fellowships.
Who can apply
A note for readers outside Brazil: this is a Brazilian federal call, and the applicant has to be a company established in Brazil. It is open to innovative companies and to small and medium startups building artificial intelligence solutions. The company puts in a counterpart of at least 20% of the project value. At least 30% of the funds are reserved for projects from the North, Northeast and Central-West regions.
The alignment the proposal has to declare
The proposal has to state its alignment with Nova Indústria Brasil (NIB, the federal industrial policy plan) and with the Plano Brasileiro de Inteligência Artificial (PBIA, Brazil’s national AI plan), the two policy frameworks the call uses as reference. Priority goes to businesses with social and environmental impact and to companies led by women. That is worth reading as part of the proposal rather than as a header formality: the call asks the project to sit inside those two plans, and the call document is what defines the terms.
A researcher inside the company is a different arrangement
Funding research at a university and paying a researcher to work inside a company are different things, even when the subject is the same. At the university the problem usually arrives formulated before the data shows up; inside the company the problem is already in production, with real data, a real deadline and an internal user asking for results. RHAE bets on the second arrangement, where the final product of the work is something the company starts using.
Where this touches engineering and geospatial data
This is my own reading, not a claim made by the call. A logistics company that processes GTFS (the General Transit Feed Specification, the standard format for public transport schedules) and needs a model to predict route delay is the kind of problem that would fit an RHAE project: real data, a real operational constraint, and a researcher dedicated to solving that specific thing. The same holds for a company working with QGIS that has a queue of manual geoprocessing tasks an AI model could shorten. It is not the only kind of company the call serves, but it is the kind this site usually discusses.
What only the call document answers
This post leaves out what the news item left out: the monthly value of each fellowship type, the total budget of the call, the result date, the submission platform, the evaluation criteria and the list of what is fundable beyond the fellowship itself. All of that is in the call document and in the FAQ PDF, on the call’s page, and that is where any interested company has to read before writing a proposal.
The deadline
Applications are open from August 19, 2026 to October 9, 2026, the cut-off date. The official CNPq announcement carries the summary of the call, and the page for call 29/2026 carries the full call document and the FAQ PDF. Both are in Portuguese.
If your company wants to talk
If your company is weighing a proposal for call 29/2026, I am available to talk about both sides of the arrangement: what the call document demands of a proposal, and what it is like in practice to have an AI researcher working inside the company, which is exactly the role the fellowship funds. Reach me on WhatsApp or at eng.dcamargo@outlook.com.br.