Roles under discussion

Pooling and security: training across companies without seeing their data.

Extend the federated training system: secure aggregation over real networks, member authentication, defences against corrupted updates that work under masking, and differential privacy where participants require it. Work with the founding engineer to take the system from single-machine tests to a multi-company pilot.

What we would look for

• —Production experience with PyTorch or a comparable framework
• —Familiarity with federated learning, secure aggregation or privacy-preserving machine learning
• —Rigour in measuring what a privacy mechanism does and does not protect

Experience with Flower or a similar federated-learning framework is an advantage.

Making the software something a company's IT department will approve.

Package the client to run inside a customer's firewall: containers, GPU support, audit logging, network-isolation checks and a clean installation process. Build the continuous-integration and reproducibility tooling behind every published result.

What we would look for

• —Experience shipping software into environments you do not control
• —Strong Linux, containers and networking fundamentals
• —C or C++ for performance-critical or system-level components is an advantage

Security-review or compliance experience (for example SOC 2) is an advantage.

Predicting which candidate materials can be synthesised.

Build the validation pipeline: structure matching for the novelty check, machine-learned interatomic potentials and DFT workflows for stability, and the synthesizability model trained on laboratory outcomes. Design the benchmarks that decide what the company can claim.

What we would look for

• —Research or industry experience in computational materials science or chemistry
• —Working knowledge of pymatgen, ASE or similar tools, and of DFT
• —Interest in solid-state synthesis and why syntheses fail

A research degree is welcome but not required; published or open-source work counts.

Responsible for everything between the product and the customer.

Run outreach to battery-materials companies and research laboratories, conduct customer conversations, and convert interest into pilots and pilots into customers. Handle incorporation and contracts with counsel, the hiring process, budgets and grants. Own the website, the company's public presence and conference work, keeping every claim within the evidence.

What we would look for

• —A technical background in chemistry, chemical engineering or materials science, or the ability to discuss R&D credibly
• —Consistent follow-through on outreach and operations
• —Experience or contacts in battery, materials or chemicals R&D is a strong advantage

Overstating a guarantee to secure a deal would misrepresent the product in a signed contract, and is not acceptable here.

Guidance on materials synthesis and solid-state electrolytes.

A few hours a month: reviewing the validation method and benchmarks, advising on what laboratories record and share, and making introductions where appropriate. Academic advisors are offered co-authorship on the methods work.

What we would look for

• —Research experience in solid-state synthesis, battery electrolytes or synthesizability prediction
• —Familiarity with how laboratories record successful and failed syntheses
• —Willingness to identify where the approach is weak

Advisory arrangements are documented, with a small equity grant.