The problem
AI agents that act without guardrails create operational, privacy and reputational risk.
Deliverables
- Agent scope and responsibilities
- Human-approval and escalation design
- Data, privacy and risk controls
What's included
- Discovery session
- Agreed scope of work
- Delivery against written deliverables
- Handover documentation
What's not included
- Third-party licences and subscriptions
- Work outside the agreed scope (change requests quoted separately)
- Ongoing support unless contracted
Delivery model
Individual Expert
Direct access to Koustov Choudhury for focused technical work.
Team
For larger or multi-disciplinary work, Koustov coordinates a vetted technical team and stays accountable for scope and quality.
Team engagements are quoted after requirements are reviewed.
Process
- Requirements
- Assessment
- Implementation
- Validation
- Delivery
- Report