The programme is the in-person facing entry point to SPDF. It is not a survey of AI tools. It is a working session in which each participant takes one signature skill — the one they would most want to scale across their organisation — and walks it through the framework with Prof. Bhat's hand on the wheel.
Decoding Your Skills into AI Agents
The SPDF Programme
Two half-day in-person sessions. Cohort of 8–12 senior managers. Afternoon project work. Each participant leaves with a v1 skill-process specification and a working v1 agent.
- → A v1 Skill-Process Specification for one chosen skill
- → A working v1 Deployable Agent Specification running on Claude Agents
- → A method they can apply, on their own, to subsequent skills
The Frame
Why tacit managerial judgement matters more in an AI age, not less. How agentic frameworks differ from the chat interfaces participants already know. A live demonstration: one manager's decision skill, already decoded, running as an agent.
Skill Elicitation
Each participant chooses one signature skill. Guided, structured reflection surfaces the heuristics, signals, and decision logic the participant actually uses. Peer challenge in pairs sharpens the articulation. Output: v1 skill-process specification.
Encoding & Context Engineering
Hands-on translation of the specification into an agent system prompt. Tool layer specification: which databases, which feeds, which internal systems the skill needs to reach. Testing against real scenarios the participant brings in.
Deployment, Review & Refinement
How the agent enters an organisational workflow without disrupting it. The review interface — how the manager monitors outputs and feeds corrections back. Scaling from one plugin to a team library. Output: v1 deployable agent specification.
Authorship & Oversight
Where human judgement must remain primary. Accountability and escalation. How the agent evolves as the manager's experience evolves. The participant's enduring role as author and steward of their embedded skill.
Throughout the programme, the BIRD financial linkages model serves as the running illustration — used so participants from any discipline can see SPDF applied end-to-end on a skill that is already well-understood.
See the worked example in detail →Cohorts run by invitation. Submitted enquiries are read personally by Prof. Bhat. Where the context fits, a discussion is arranged before the participant is admitted. Programme fees are shared with admitted participants.