IES's Second Tranche Is Live: Twelve $500,000 AI Seedlings and a $3.75 Million Bet on Teacher Preparation, Both Due December 11

One competition gives you a year and $500,000 and warns you that "many teams will not progress past Phase One." The other gives you up to three years and $3.75 million, and makes Years 2 and 3 contingent on recruiting a cost-share partner you don't need to have yet. The Institute of Education Sciences posted both on October 16, 2026. Together they are the second of the agency's three FY2027 grantmaking tranches.

When IES restarted research grantmaking in August after roughly fifteen months of silence, it previewed this tranche as a late-September or early-October release. It arrived a couple of weeks late, with deadlines that land in the middle of the academic fall:

Both are open to any institution "that ha[s] the ability and capacity to conduct scientifically valid research." That explicitly includes for-profit organizations, not just universities and nonprofits. Both RFAs are posted on IES's site (305J, 324T).

From Seedlings to Scale: An AI Systems Bet, One Year at a Time

S2S runs on a venture-style three-phase structure under IES's Accelerate, Transform, Scale (ATS) initiative. Phase One is the "seedling": one year to understand the problem, prototype a solution, and argue for more money. Phase Two adds two years of development and testing in a limited setting, and teams are selected by peer review of their Phase One results. Phase Three, scaling, is at IES's discretion.

For FY2027, Phase One is restricted to a single focus area: C-STEP — Coordinating Systems to Translate Evidence into Policy and Practice. IES wants "innovative AI- and technology-enabled solutions" that help school, district, and state leaders work out which factors are actually driving poor outcomes, choose evidence-based programs that fit local conditions, estimate the full cost of implementation, and model likely effects before committing. The RFA's examples include knowledge networks linking evidence to costs, system dashboards, tools that surface unintended consequences, and modeling platforms for comparing implementation strategies. IES says it wants ideas "inspired by successful technological solutions from other fields," naming healthcare and business.

A narrow reading would be "build a better evidence clearinghouse." The RFA pushes against that. Strong proposals "will address the root causes of implementation failure rather than focus on a single proximal cause." The example given is a district fixing curriculum quality while chronic absenteeism quietly caps the gains. Reviewers score breakthrough potential first. The questions are whether the problem, if solved, would change leaders' ability to act on evidence; whether the theory of change is credible; and whether the applicant names the assumptions Phase One must test.

Phase One has three required deliverables:

1. A Proof-of-Concept Report covering user needs, market opportunity, solution-market fit, and build requirements
2. An Early-Stage Prototype, presented at an ATS Hub demo day in Washington in early 2028
3. An R&D Plan for Phase Two, including timeline, staffing, and partner commitments

Teams must also attend ATS Hub capacity-building, including a one-day workshop in late summer or early fall 2027, and use the Hub's Learn to Scale Toolkit. The RFA notes that Phase One and Phase Two may need different expertise. A university team that plans to be the research partner rather than the builder should say so and name who builds.

Who should apply: interdisciplinary teams that pair education-systems researchers with people who have shipped data or decision-support products, and teams with a state or district partner willing to be the first user. Edtech firms can be the prime, which makes S2S a realistic non-SBIR route for companies whose product is a district-level analytics or planning tool. Who should not: teams with a finished product looking for an efficacy trial. Phase One explicitly does not expect a functioning solution, and reviewers will read a mature product as a mismatch.

Transformative Research in Special Education: Fixing the Pipeline, Not the Classroom

NCSER's competition is aimed at preservice teacher education systems. Applicants must "dramatically improve" how those systems attract, train, and retain preK–12 general and special education teachers who can serve students with or at risk for disabilities.

The RFA's problem statement covers the pipeline: falling enrollment and completion in preparation programs, certified special educators who never take special education jobs, wide variation in quality across a growing number of alternative-certification routes, and general-education candidates who get "very little training" for teaching students with disabilities, even though many of those students spend much of their day in general classrooms.

Requirements worth noting:

• The focus must be preservice. In-service teachers can be in the sample where appropriate, but the research focuses on candidates and new teachers.
• Teacher outcomes are measured in Phase 1. Student outcomes wait for Phase 2. Applications still need an explicit theory of change running from system change to teacher outcomes to student outcomes.
• Project-specific milestones. Applicants propose their own Phase 1 deliverables and metrics. IES uses them to judge whether the project is on track and whether it qualifies for Phase 2, which could add up to $3 million over two more years.
• Co-design partners. The application must describe a plan for co-design partners: teacher candidates, in-service teachers, education deans, teacher educators, principals, parents, and students with disabilities. They don't have to be in place at submission, but continuation funding for Years 2 and 3 depends on having them before the end of Year 1, with letters documenting roles.
• 5 percent cost share, secured in Year 1. On a $3.75 million request that is $187,500 in cash or in-kind contributions, from state, local, or private sources. As with co-design partners, continuation depends on a signed cost-share agreement by the end of Year 1, submitted with the first annual report.

Those last two provisions shape strategy. A team without a committed state agency, district consortium, or foundation partner can still apply. But it is effectively borrowing against a partnership it has to deliver within twelve months, and the downside is losing two-thirds of the award. The strongest applicants will arrive with at least a letter of interest from a state education agency or educator-preparation network that controls the system being changed. Those are the partners best placed to provide in-kind cost share, such as data access, staff time, and clinical placements.

Rules That Apply to Both

Several FY2027 changes apply across both RFAs:

• Appendix F: Alignment with SEER 27 is new and required. Applicants explain how the project meets the 2027 Standards for Excellence in Education Research across four domains: objectivity and scientific integrity, relevance, open science, and intervention research. Program officers review it before award.
• Research security compliance is a pre-award condition.
• AI-assisted writing is allowed, but the authorized representative certifies the application as "true, complete, and accurate" regardless of how it was drafted.
• Persistent identifiers for all key personnel are required as a condition of award.
• Page limits are enforced by deletion. IES removes pages past the eighth (305J) or eighteenth (324T) narrative page before review.

IES applies an institution's federally negotiated indirect rate. Neither RFA imposes a cap below it.

Timing and the December Pile-Up

The third tranche is the one most of the field is waiting for. It includes the core field-initiated Education Research Grants (84.305A) and Special Education Research Grants (84.324A), which IES said in August would open in December 2026. Applicants are deciding now whether to spend November on a speculative S2S or Transformative proposal or save effort for the larger, more familiar competitions.

Here is our read. S2S is a good bet for a small, fast, product-minded team with a district partner. Twelve slots at $500,000 is a real competition, and a one-year deliverable plan is easier to write than a five-year efficacy design. The Transformative competition fits best for a teacher-preparation researcher whose work is already system-level and who can line up a state partner by spring. For a classroom-intervention researcher, 84.324A in the third tranche is the better fit.

Submit the letter of intent by October 29 even if you are not sure. It is non-binding, and IES uses it to recruit reviewers with the right expertise, which matters for an AI-systems competition where the ideal panel looks different from a typical education-research study section. Then register in Grants.gov, confirm SAM.gov, and plan backward from December 11.

Whether you are writing an eight-page seedling pitch or an eighteen-page systems-change design, Granted can help you turn the RFA's review criteria into a narrative outline before the letter-of-intent date arrives.