Africa will not follow the materials frontier. Africa will help define it.
Build open, interoperable, AI-native tools that accelerate materials discovery for African priorities.
01
Africa-first · End-to-end · Open science
MaterialsGPT Africa
An African Grand Challenge for LLM-Powered Materials Design and Discovery. Build the AI stack for Africa’s materials future.
Materials design · LLM agents · Simulation · Lab planning · Open data · Africa-MRS showcase · Sponsor-ready science · Materials design · LLM agents · Simulation · Lab planning · Open data · Africa-MRS showcase · Sponsor-ready science ·
Sponsor front page
Back the first African build effort for AI-native materials discovery.
Sponsors support a visible, high-quality, Africa-led scientific campaign that moves beyond talks and toward tools: LLM applications, open data rails, integrated discovery workflows, and prototypes that can be showcased to the Africa-MRS community.
Why this is sponsor-ready
MaterialsGPT Africa creates a credible bridge between AI, materials science, African industrial priorities, and continental scientific visibility. It gives sponsors a concrete way to support talent, open infrastructure, and end-to-end discovery capacity.
Visible continental platformWinning teams showcase at Africa-MRS Kenya 2026, placing sponsor-supported work before a specialist materials research audience.
Real technical outputsThe challenge emphasizes working tools, documented methods, open interfaces, and validation pathways rather than slide-only ideas.
Strategic African relevanceProjects are anchored in energy, water, infrastructure, green industry, electronics, and mineral value addition.
Why this matters
Africa should not wait for the materials frontier to arrive.
Materials innovation is the hidden engine behind clean energy, resilient infrastructure, advanced manufacturing, water security, agriculture, and mineral value addition. This challenge frames Africa as a builder of the next generation of scientific infrastructure.
Not generic AI
Every prototype must connect an LLM or AI system to a real materials discovery decision.
02
Not isolated demos
Teams must show pipeline position, integration logic, and a path toward validation.
03
African missions
Concrete materials problems, not abstract AI.
The challenge is grounded in domains that matter for Africa’s industrial, climate, and scientific future.
Energy storageBatteries and storage materials for African energy systems.
Green industryCatalysts for fertilizers, fuels, and local manufacturing.
InfrastructureLow-carbon cement and durable construction materials.
Water securityFiltration, membranes, and remediation.
ElectronicsSolar, semiconductor, and electronic materials.
Mineral value additionMaterials for beneficiation and advanced processing.
End-to-end frame
From question to candidate to validation pathway.
Each team must locate its prototype within the full discovery lifecycle and explain upstream inputs, downstream outputs, and validation route.
KnowledgeExtract evidence from literature, patents, datasets, and lab notes.
DesignGenerate candidate compounds, recipes, structures, and hypotheses.
PredictEstimate properties using ML, simulation, or surrogate models.
PlanTurn predictions into resource-aware experimental decisions.
ValidateDefine benchmarks, lab pathways, and deployment logic.
Challenge-first model
Start with Africa's hardest materials problems.
MaterialsGPT Africa begins before teams write code. Researchers, labs, industry partners, and institutions propose concrete challenge problems. Teams then form around the strongest missions or bring their own approved challenge.
Challenge - Teams - Systems
The program is designed to avoid disconnected demos. Problems come first, teams organize around them, and every prototype must show where it plugs into the end-to-end discovery pipeline.
Grand rule: a team can either adopt a published challenge, form around a proposed challenge, or submit its own Africa-grounded challenge for approval.
Challenge ownersPeople who propose challenges can mentor, advise, or help define success criteria.
Challenge bankApproved problems are published for participants to explore and join.
Team formationParticipants sign up by skills, track preference, and challenge interest.
Pipeline requirementEach team must explain upstream inputs, downstream outputs, and integration potential.
Propose a challenge
Submit a materials problem with African context, desired outcome, and preferred track.
Curate and publish
Organizers align approved challenges to the six tracks and publish them as live missions.
Join or form a team
Participants adopt a challenge, build a team, or propose their own problem for approval.
Grand challenge tracks
Six tracks, one African discovery stack.
The tracks are designed as interoperable modules. Teams may specialize, but they must show how their tools connect to the wider pipeline.
Turn Africa’s materials knowledge into machine-actionable intelligence.
Build LLM systems that read papers, reports, datasets, patents, or lab notes and convert them into structured facts, hypotheses, material-property relationships, and research maps.
Challenge: Create a scientific knowledge engine that helps an African researcher move from scattered literature to a testable materials hypothesis.
Design candidate materials before they enter the lab.
Use LLMs and generative AI to propose new compounds, compositions, structures, or recipes optimized for batteries, catalysts, membranes, cement, electronics, or other African missions.
Challenge: Build a materials design copilot that can propose, explain, and rank candidates for a target African application.
Make simulation faster, cheaper, and easier to use.
Build tools that connect LLMs to physics-based simulation, surrogate models, property predictors, or workflow automation so researchers can rapidly test promising candidates.
Challenge: Create an AI assistant that helps move from a proposed material to predicted properties, simulation inputs, and interpretable results.
Let AI plan the next best experiment.
Develop LLM-guided experimental planning systems that recommend synthesis routes, characterization plans, safety checks, parameter sweeps, or active-learning loops.
Challenge: Build an experiment planner that turns model outputs into practical, resource-aware laboratory decisions.
Build the data rails for African materials discovery.
Create open data tools, schemas, APIs, validation layers, benchmarks, or repositories that make African materials data easier to share, compare, reuse, and trust.
Challenge: Design the infrastructure layer that allows many teams, labs, and countries to contribute to one discovery ecosystem.
Connect the pieces into an end-to-end discovery workflow.
The flagship systems track. Teams combine knowledge extraction, design, prediction, planning, and validation logic into a coherent discovery workflow.
Challenge: Build a working pipeline that shows how an African materials problem can move from question to candidate to validation pathway.
Hackathon format
A mission program, not a weekend coding contest.
The format is deliberately gamified and systems-oriented: teams compete, collaborate, integrate, and prove that their tools can contribute to Africa's materials discovery stack.
01
Mission scoping
Teams select or propose a challenge, define the materials target, and map where their system sits in the discovery pipeline.
02
Pipeline design clinic
Mentors review upstream inputs, downstream outputs, LLM strategy, validation pathway, and interoperability assumptions.
03
Build sprint
Teams develop prototypes with checkpoints on scientific rigor, AI usefulness, data quality, usability, and integration readiness.
04
Integration review
Teams demonstrate whether another team, lab, or platform could actually use their outputs in a wider discovery workflow.
Working systemA prototype that performs a real discovery task.
Pipeline diagramA map of inputs, outputs, users, and downstream value.
Interface noteAPI, schema, file format, or workflow handoff.
Mission pitchA concise case for Africa relevance and scientific impact.
Challenge timeline
From challenge call to Africa-MRS showcase.
The timeline turns the hackathon into a campaign: first define the problems, then form teams, then build systems, then judge and showcase the strongest outputs at Africa-MRS Kenya 2026.
01 | Call for Participation - by September 2026Open call issued across African universities, research institutes, and laboratories. Applicants submit team composition, project concept, and technical approach. Shortlisted teams are notified and admitted to the preparation phase.
02 | Team Formation & Preparation - 5 to 16 October 2026Two-week preparation phase. Participants refine team composition and sharpen project ideas through introductory workshops on LLMs for scientific discovery, materials informatics, and open science practices.
Infrastructure setupTeams set up collaborative infrastructure: GitHub repositories, Slack or Discord channels, shared environments, notebooks, and coordination boards.
03 | Hackathon Weekend - 23rd to 26th October 2026Friday 23 October 2026, 14:00 CAT to Monday 26 October 2026, 12:00 noon CAT. An intensive collaborative sprint to develop functional prototypes and research tools.
Build and mentor supportTeams develop working prototypes, research tools, AI models, and demonstration systems. Mentors provide real-time technical and scientific guidance with intermediate check-ins.
04 | Final Presentations & Judging - 30 October 2026Teams present to expert judges including materials scientists, AI researchers, and industry partners. Winners are announced and prizes awarded.
Judging focusScientific relevance, technical quality, innovation, pipeline integration, open science adherence, and impact potential.
05 | AMRS 2026 ShowcaseWinning teams showcase in Kenya at Africa-MRS, 5 to 10 December 2026.
Evaluation and gamification
Reward systems that can move science.
Judging is designed to reward rigorous science, meaningful AI, end-to-end integration, Africa relevance, and reusable infrastructure. Teams earn bonus recognition for collaboration across tracks.
Scientific relevance20%
AI/LLM innovation20%
End-to-end integration25%
Africa impact20%
Openness and interoperability15%
BonusCross-team integration
BonusMulti-track pipeline linkage
BonusReal-world validation feasibility
Best End-to-End Discovery SystemFor the team that best connects knowledge, design, prediction, and validation.
Best Africa Impact SolutionFor the strongest response to a clearly articulated African materials challenge.
Best Scientific LLMFor the most useful LLM application in reasoning, extraction, design, or workflow automation.
Best Data Backbone ContributionFor reusable schemas, APIs, datasets, or infrastructure that other teams can build on.
Prize pool
Reward teams that can build, explain, and scale.
Prize cards are designed to be sponsor-friendly and clear for participants. Assumes five members per winning team.
🥇
First place
Grand Challenge Winner
USD 0000 each
Total team award: USD 0000
Conference travel support, presentation slot, and AMRS showcase pathway.
Conference travel support, presentation slot, and AMRS showcase pathway.
🥈
Second place
Pipeline Builder Award
USD 0000 each
Total team award: USD 0000
Scientific visibility, network access, and continued mentorship.
Scientific visibility, network access, and continued mentorship.
🥉
Third place
Open Science Award
USD 0000 each
Total team award: USD 0000
Recognition, open science community inclusion, and development feedback.
Recognition, open science community inclusion, and development feedback.
Open science and research infrastructure
Build in public, so Africa keeps the infrastructure.
The hackathon promotes open science principles and collaborative research practices. Each team contributes to a shared African materials informatics ecosystem, not a closed one-off output.
Open outputs are required.
Participants use open-source tools, publish code and documentation, follow FAIR data principles, and avoid closed outputs.
OS
Public GitHub repository
Each team maintains a public repository containing documentation, code and models, datasets where applicable, and demonstration materials.
GH
Collaboration stack
Teams use GitHub, Slack or Discord, Notion or Trello, Google Colab or JupyterHub, and Zoom or equivalent.
FAIR
Eligibility
Who should participate?
MaterialsGPT Africa is designed for multidisciplinary teams that combine materials science, AI, software, data, and scientific domain expertise.
Materials and physical sciences
Materials scientists, chemists, physicists, and computational scientists working on real-world discovery problems.
01AI, software, and data
AI/ML researchers, engineers, software developers, and data engineers building intelligent scientific systems and tools.
02Emerging interdisciplinary talent
Graduate students, early-career researchers, and interdisciplinary teams committed to scientific innovation.
03Teams are strongly encouraged to be multidisciplinary.
Organising committee
The people stewarding MaterialsGPT Africa.
The committee brings together African and international expertise across materials science, electrochemistry, computation, research infrastructure, and Africa-MRS coordination.
International Organising Committee
Prof Cecil NM OumaChairAIMS Research & Innovation Centre (AIMS RIC)
Dr Emily AradiComputational Materials ScienceUniversity of Manchester
Dr Joyce ElisadikiElectrochemistry & Energy MaterialsUniversity of Dodoma, Tanzania
Dr Pamela KilaviMaterials InformaticsStrathmore University, Kenya
Dr Edwin MapashaCondensed Matter PhysicsUniversity of Pretoria
Prof Ponnadurai RamasamiComputational ChemistryUniversity of Mauritius, Mauritius
Local Organising Committee
Prof Robinson MusembiSemiconductor PhysicsUniversity of Nairobi
Dr Mike AtamboApplied PhysicsTechnical University of Kenya
Dr Francis GaithoCondensed MatterMasinde Muliro University
Dr Winfred MulwaMaterials ChemistryEgerton University
Africa-MRS Representative
Dr Victor OdariAfrica-MRSMasinde Muliro University
Dr. Erick NjoguAfrica-MRSMasinde Muliro University
Sis. Dr. Mary TaabuAfrica-MRSThe University of Nairobi
Technical Partner
Prof Paul Ndirangu KioniAI & Research InfrastructureOrigin Research, Nairobi
Dr Eng. Anthony GithinjiSemiconductor Technologies LimitedDeKUT, Nyeri
Prof Francis NyongesaMaterials ScientistUniversity of Nairobi
Dr Katawoura BeltakoKnowledge and Technology TransferAIMS RIC
Walelign SewunetieKnowledge and Technology TransferAIMS RIC
Call for participation
Apply, propose, coach, or form a team.
Participants can enter by proposing a challenge problem, volunteering to coach a challenge-linked team, or applying to join/form a team around an approved challenge. Coaches are tied to challenge problems and teams, not assigned as floating advisors. Teams may also bring their own Africa-grounded problem if it is approved by the organizers.
Strong entries should show
- A concrete African materials problem.
- A clear LLM or AI role in the discovery pipeline.
- Practical handoff to design, prediction, validation, or data infrastructure.
- A multidisciplinary team, coaching role, or a plan to form one.
Linked applications
Challenge proposals feed team formation.
The two forms are linked through Firebase Firestore. Challenge proposals enter the Challenge Bank as pending records. Once organizers mark a challenge as approved, it automatically appears in the live challenge list and in the team application dropdown.
Winner showcase
Winning teams will showcase at Africa-MRS Kenya 2026.
The hackathon winner will be positioned to present their work at the 13th International Conference of the African Materials Research Society in Kenya, creating a bridge from prototype to continental scientific visibility.
Organized with Africa-MRS as a scientific partner.
Linked to AMDA as a longer-term pathway for validation, mentoring, and ecosystem building.
Built for sponsors who want credible African science, talent, and infrastructure impact.