Three pillars. One lifecycle. Every engagement runs from strategy through production to embedded capability — with your team owning the result. Physical AI for the machines that move the real world: robots · factories · vehicles · aircraft · critical infrastructure. Sovereign-first: Mistral open-weight → on-prem → edge.
Strategy, governance & embedded leadership for physical systems — assess where you stand across robotics, industrial, automotive, aerospace, energy, and logistics verticals; identify the highest-impact use case; embed fractional executive AI leadership; and ensure EU AI Act + functional-safety readiness before committing engineering resources.
Two weeks. A written readiness report ranked by severity — covering OT/IT boundary, sensor-data pipelines, edge-inference readiness, and safety-regime maturity — plus a 90-day roadmap you can defend to your board
Four weeks to a strategy document, a business case, an ROI model, and a 12-month execution plan — scoped for industrial operators where OEE, safety-regime timelines, and physical-system procurement cycles are the real constraints, not just board optics
Twelve to twenty-four weeks to risk-classify your AI systems, complete the conformity assessment, produce the Annex IV technical documentation, and stand up post-market monitoring — with special depth on Annex III high-risk categories for autonomous, industrial, and robotics deployments
Part-time executive AI leadership — six to twelve months, two days a week — owning the AI P&L, governing physical-AI risk, chairing the AI-safety board, aligning OT/IT AI policy with standards, and building the regulator-facing narrative for industrial, automotive, and energy organisations
Engineer & ship — the differentiator. Architect, build, test, and deploy Physical AI systems to production: robots and autonomous systems, industrial manufacturing, automotive and mobility, aerospace and aeronautics, energy and critical infrastructure, logistics and supply-chain. Sovereign-first by default.
Build engagements implement the Mistral open-weight stack: fine-tuning on your engineering data (Mistral Forge), agentic workflows (Mistral Studio), and on-prem / air-gapped inference at the edge (Mistral Compute pattern) — so the model stays on your infrastructure, not a cloud vendor's.
Eight weeks. A fine-tuned open-weight model — Mistral, Llama 3, or Qwen — trained on your proprietary industrial data (maintenance manuals, MES/PLC logs, technical documentation) and running on infrastructure you control, not a frontier API you rent
Twelve weeks to a production-grade multi-agent system that serves as the software and control-plane complement to your cyber-physical stack — fleet intelligence, SCADA-adjacent orchestration, or autonomous operations — with the eval harness, the observability stack, and the SRE handoff your team needs to operate it
Sixteen weeks to AI running on the edge — inside a robotics cell, an AGV/AMR fleet, an ADAS or AD stack, a UAS/drone autonomy system, or a substation — with the safety evidence, the SRE handoff, and the integration your operations team will accept
Twelve weeks to harden an edge or embedded AI pilot stuck before production — on constrained hardware, inside safety envelopes, under latency and reliability requirements the pilot was never designed to meet
Turn cameras, LiDAR, radar, and IMUs into one trustworthy world-model — the Sense layer of the Physical AI Stack, engineered for the edge and degraded conditions.
Functional-safety evidence for AI in safety-critical machines — HARA, ASIL/DAL/SIL decomposition, and assurance cases mapped to ISO 26262, DO-178C, and IEC 61508. A notified body certifies; I engineer the evidence they assess.
Capability & embed — transfer the methodology to your team. Physical AI engineering upskilling and the agent-augmented L0→L3 delivery methodology that turn pilots into a permanent organisational capability across your industrial verticals.
Book a discovery call. No pitch, no pressure — just an honest conversation about your situation and whether I can help.