Which dependencies, APIs, or operational risks are limiting delivery and reliability?
What we need
Service inventory, selected code and API documentation, incident evidence, dependency maps, and stakeholder interviews. Credentials and personal data are not needed for initial scoping.
Scope boundaries
A diagnostic is not a penetration test or implementation engagement. Load testing and code changes require explicit scope.
Is a simulation-trained policy appropriate for this robot, task, and operating domain?
What we need
Robot specifications, task requirements, available simulation assets, control interfaces, success metrics, and physical safety constraints.
Scope boundaries
Factory and household use cases remain separate programs. Training, hardware trials, and safety qualification are not included unless explicitly agreed.
Outputs to agree
Domain-specific feasibility assessment
Simulation and reward-design requirements
Evaluation and sim-to-real risk plan
Training and deployment options, including when not to proceed
Duration is an indicative working window, subject to access and stakeholder availability. Each engagement is quoted after scoping; the signed SOW determines the fee, outputs, exclusions, milestones, and acceptance criteria. Implementation and ongoing support are separate decisions.
Start with Fitzroy
Define the next decision before the next build.
Discuss the system question and the information available. Scope, fee, timing, and contractual terms are agreed before kickoff.
