MIDAS African Faculty Fellowship 2027: $35,000 Stipend, Housing and $50,000 Research Pilot Grant

MIDAS African Faculty Fellowship 2027: $35,000 Stipend, Housing and $50,000 Research Pilot Grant

The Michigan Institute for Data Science (MIDAS) at the University of Michigan is preparing the 2027 cohort of its African Faculty Fellowship. The programme is for current faculty members at African universities, and for researchers at African research organisations whose roles are comparable to university faculty positions. It supports a two-year fellowship built around one year in residence at the University of Michigan in Ann Arbor and a return to the fellow’s home institution in Africa.

This is a particularly substantial opportunity for a researcher who has strong domain expertise but wants to develop serious AI methods for science or engineering. The official programme describes AI broadly: machine learning, robotics, Bayesian inference and simulation are all within its stated meaning. The scientific areas include biological and biomedical science, earth and environmental science, engineering, physical science and mathematics, subject to the programme’s exclusions.

The financial support is more than a salary line. Fellows receive an annual $35,000 stipend, medical insurance during their U-M residence, housing, visa support and travel between U-M and Africa. The programme also supports computers and software, conference travel, training workshops and publication fees. After the fellow returns to Africa, the programme provides a $50,000 pilot grant for the fellow’s training, research and impact plan. The fellowship also sets aside $40,000 for the U-M faculty host’s joint research project, which helps make the collaboration practical without treating that host budget as personal funding for the fellow.

The application portal for the 2027 cohort is scheduled to open on August 17, 2026. Phase 1 applications are due by 11:59 p.m. EDT on October 5, 2026. Applicants who advance then submit Phase 2 materials by 11:59 p.m. EDT on November 23, 2026. If your work sits at the intersection of an important African research problem, a credible AI learning plan and a clear institutional impact pathway, this is a programme worth preparing for now.

Key details

The dates and amounts above come from the official 2027 application information. The application portal is not scheduled to open until August 17, so applicants should use the current programme page to prepare and then confirm the live portal instructions when it opens.

What the fellowship actually offers

The fellowship is designed as a cycle rather than a one-way visiting appointment. During the U-M residence, the fellow works with a U-M faculty host and the MIDAS community, develops AI skills and carries out a joint research project. The fellow is expected to spend at least 50% of the in-person residence time at the MIDAS postdoctoral offices, in addition to working within the host’s research group.

The programme provides the fellow with a $35,000 annual stipend and medical insurance for the U-M residence. It also says it will provide housing and cover visa fees and travel costs between U-M and Africa. Separate programme funds can cover a computer and software licences, research conference travel, training workshop fees and publication costs. These details matter because a candidate should assess the fellowship as a supported relocation and research period, not simply compare the stipend with a local salary.

The second part happens after the fellow returns to the home institution in Africa. The $50,000 pilot grant is intended to carry out the training, research and impact plan developed during the fellowship. The plan might include an AI-in-science course at the home university, a regional workshop, exchange activity with U-M researchers or a research project that can grow into a larger collaboration. The official examples are broad, but the common thread is local capacity and a credible route from the U-M experience to work that remains useful in Africa.

The U-M host receives $40,000 for one joint research project with the fellow during the residence. The page explains that the host budget may cover a limited amount of host effort, collaborating students or other approved project costs. Applicants should understand the distinction: this is collaboration infrastructure, not a second stipend paid to the applicant.

Who is eligible

The first eligibility test is institutional. You must have a primary appointment as a faculty member at a university in Africa, or work at an African research organisation in a role that carries out research in a capacity similar to university faculty. This is not a general scholarship for anyone who lives in Africa, and it is not aimed at applicants whose only connection to the continent is the topic of their research.

You must hold a PhD before the fellowship start date of July 1, 2027. The programme is aimed at early- and mid-career researchers. To meet its timing rule, either your PhD conferral date or the start of your first faculty-level appointment, whichever came earlier, must be on or after July 1, 2017. A PhD or first faculty appointment dated June 30, 2017 or earlier is outside the stated window.

Your PhD may be in any field. The field of the proposed work is what must fit the programme’s scientific scope. That makes the opportunity broader than a computer-science fellowship: a researcher can bring expertise in a domain such as environmental systems, physics, engineering or basic biology and propose a serious AI direction without needing a PhD in AI.

The proposed research must use AI methodologies to address a significant question in science or engineering. Biological and biomedical work is allowed when it focuses on basic biological processes or disease mechanisms and propagation, including in-silico, in-vitro or animal-model work. Proposals focused on developing or improving disease diagnosis, prevention or treatment are explicitly excluded. A clinical AI product pitch is therefore a poor fit even if it uses sophisticated machine learning.

The research areas and the fit test

The official page lists four broad areas: biological and biomedical science, earth and environmental science, engineering, and physical science and mathematics. The list is intentionally wide, but the fellowship is not asking for an AI method in isolation. The candidate is expected to show a meaningful domain challenge and a reason AI could enable a research advance.

A good fit has three connected parts. First, identify a domain question that matters scientifically and, ideally, has importance for Africa or the applicant’s region. Second, explain which AI methods the applicant needs to learn or strengthen, such as a particular modelling, inference, simulation or robotics approach. Third, show how the work will continue after the U-M residence through the applicant’s home institution, collaborators, students or research community.

The programme calls the required document an Individualized Training, Research, and Impact Plan, or I-TRIP. That name is useful: the document is not only a research proposal. It must show what the candidate will learn, what research they will do with a U-M host and what capacity or impact they will build after returning home.

How the two-phase application works

Phase 1 is the gateway. Applicants submit a CV and an I-TRIP of up to four pages, using a font size of at least 10. The page says the research vision and plausible approach may occupy only a page or two of that total. MIDAS is not asking for a complete grant proposal with every technical detail, a full related-work review and a comprehensive bibliography. A high-level description of the main ideas is enough at this stage, but it still needs to be coherent and credible.

The I-TRIP must include a research vision, an approach for achieving it, a plan for developing AI-in-science capacity after returning to Africa, and the qualifications or experiences that make the candidate a strong fit. A short list of essential citations may be included. The capacity plan could involve teaching, workshops, regional events, exchanges, collaboration with local research networks or another feasible activity that matches the institution’s actual resources.

An applicant may optionally include a letter of support from an eligible U-M faculty host in Phase 1. Finding a host before applying is not required. If an eligible host supports the Phase 1 application, the application automatically advances to Phase 2. Other eligible Phase 1 applications are reviewed by U-M faculty who have expressed interest in hosting; applications that receive host support move forward.

In Phase 2, the candidate works with the U-M host to improve or revise the I-TRIP. The supplements include the CV, up to three AI keywords, a research tagline of no more than eight words, and a third-person research blurb of no more than 300 words for a general audience. The candidate submits a revised I-TRIP, three best papers or other written works where they were a primary contributor, a host letter of support, a home-institution letter approving leave and confirming that the candidate will resume the position, and contact information for two or three references.

The host letter is different from a reference letter. It should explain the candidate’s qualifications, the importance of the research, the fit with MIDAS and how the host will collaborate on the research and impact plan. References are contacted through the application system, and the official page recommends sending those requests early.

Timeline for the 2027 cohort

The portal is scheduled to open on August 17, 2026. Phase 1 is due October 5, 2026, at 11:59 p.m. EDT. The official page says all applications received by that deadline will receive full consideration. Phase 2 is due November 23, 2026, at 11:59 p.m. EDT.

Acceptance notifications start in early February 2027. Selected applicants have two weeks to decide whether to accept the offer. The expected start date is July 1, 2027, although the programme says special cases may be negotiable.

That schedule creates a useful preparation sequence. Before the portal opens, settle the domain question, map the AI training gap and identify the home-institution outcome you want to build. In August and September, draft the four-page I-TRIP and review the prospective-host list. In September, ask possible referees whether they can respond and begin the institutional conversation about leave and return. After Phase 1, keep time available for host conversations and a fast revision cycle before the November deadline.

How to prepare a stronger I-TRIP

Start with the scientific problem, not with a list of fashionable AI tools. A reviewer should be able to state the research question after reading the opening paragraph. Then explain why the existing approach is insufficient and what a new AI method might make possible. Avoid promising a complete solution to a broad national problem in four pages. A narrow, testable research direction is easier to assess and easier to connect to a realistic training plan.

Make the training need concrete. If you need probabilistic modelling, simulation, robotics or machine learning for a specific kind of data, say so. Explain what you already know and what you expect to learn at U-M. The selection criteria specifically ask about the candidate’s potential to acquire strong AI skills and whether the candidate and host have thought carefully about skill development.

Treat the impact plan as an operational plan. Name the people who would benefit, the activity you could run after returning, the resources it would require and the first evidence that it worked. A proposed graduate course, regional workshop or collaboration should have a plausible audience and delivery path. The plan does not need to be large; it needs to be feasible for the home institution and useful to the local or regional research community.

Use evidence efficiently. Choose papers that show your primary contribution, not merely the prestige of a venue. In the I-TRIP, connect prior work to the proposed direction and explain the specific expertise you bring. If a prospective host has given feedback, use it to make the collaboration more specific, while retaining ownership of the research vision.

Common mistakes to avoid

The most serious mistake is confusing an eligible AI-in-science project with a healthcare application. A proposal centered on diagnosis, prevention or treatment is expressly outside scope. If the work is biomedical, describe the basic biological process or disease mechanism being studied and make that boundary clear.

Another mistake is treating the I-TRIP like a conventional long grant proposal. Four pages do not allow an exhaustive literature review. Spending most of the document on background leaves too little room for the training and return-home plans that define this fellowship.

Do not assume that any U-M faculty member can host you. The programme says the host must be a U-M tenure-track faculty member, research scientist or research faculty member. Your prior faculty mentor, host or principal investigator is ineligible to act as the host, and a U-M host can support at most one research fellow across the relevant MIDAS fellowships. Check the prospective-host list and eligibility rules before asking for support.

Do not leave institutional leave approval until Phase 2. The home institution letter must approve the leave and affirm that you will resume your position after the programme. That is a substantive commitment, not an administrative upload. Begin that conversation early enough for the responsible office or dean to review it.

Finally, do not compress references into the final hours. The portal lets applicants send reference requests before submitting the final application. Use that feature early and give referees the I-TRIP, CV and a short explanation of the fellowship so their letters address the relevant research, training and impact questions.

Frequently asked questions

Do I need a U-M host before Phase 1?

No. The official application page says identifying a U-M faculty host in advance is not required. A host letter is optional in Phase 1, although an eligible application that includes one automatically passes to Phase 2. Without one, an eligible application may be reviewed by U-M faculty interested in hosting.

Is this only for computer scientists?

No. The candidate’s PhD can be in any field. The proposed work must fall within the listed science and engineering domains and use AI methods, but the programme is intended to bring domain researchers into AI in science.

How much direct funding does a fellow receive?

The programme states an annual stipend of $35,000, plus medical insurance for the U-M residence. It also provides housing, visa and travel support, and funding for specified research and training costs. After the fellow returns to Africa, the programme provides a $50,000 pilot grant for the training, research and impact plan. The $40,000 host-project allocation is administered for the joint research project and should not be counted as personal stipend.

Can my proposal be about medical AI?

Only within the stated boundary. Basic biological or biomedical research, including work on disease mechanisms and propagation, can fit. A proposal to develop or improve disease diagnosis, prevention or treatment is excluded according to the official page.

When does the 2027 fellowship begin?

The expected start date is July 1, 2027, though the programme says that special cases may be negotiable. The two-year structure includes one year in residence at U-M and continued work through the home institution in Africa.

What happens after Phase 1?

Applicants who advance collaborate with a U-M host to revise the I-TRIP and submit the Phase 2 supplements, including three best papers, host and home-institution letters, a general-audience research blurb and two or three reference contacts. Phase 2 is due November 23, 2026.

Official links and next steps

Start with the MIDAS African Faculty Fellowship application page. It contains the programme scope, eligibility rules, prospective-host information, application package and selection criteria. The same page states that the 2027 portal will open August 17, 2026.

Before the portal opens, confirm your PhD and first-faculty-appointment dates, check that your home institution is eligible, and sketch a four-page I-TRIP around one research question. Then review the prospective-host list, identify the AI skills the project needs and start the leave conversation at your institution. The first hard deadline is October 5, 2026; the second is November 23, 2026. Recheck the official page when the portal opens because the live submission instructions are the final authority.