Speaking
Clear thinking on AI you can prove.
Practical, evidence-led talks and workshops for leaders and engineers on evaluating, securing and governing AI systems.
Talk topics
What I speak about.
Release gates for LLM and agentic systems
How to turn groundedness, safety and consistency into thresholds a release has to clear.
Red-teaming AI agents: prompt injection, tool poisoning and excessive agency
How agents are attacked in practice, and how to test and guard against it using OWASP LLM Top 10 and MITRE ATLAS.
AI governance in the GCC that speeds decisions up
Risk tiering, model inventories and evidence trails aligned to NIST AI RMF and UAE IA / NESA and ADHICS, designed to speed decisions up.
Reliability engineering for autonomous agents in production
Failure modes, fallbacks, SLOs and observability for agents that act on their own.
Each topic works as a keynote, an executive briefing or a hands-on workshop. Details of recent talks are available on request.
Formats
For the room you have.
- Keynotes
- Conference and summit talks on trustworthy, provable AI.
- Executive briefings
- Board and leadership sessions on AI risk, assurance and governance.
- Workshops
- Hands-on sessions on evaluation pipelines and red-teaming for engineering teams.
Press kit
Bio and headshot.
- Short bio
- Satya Prakash Solanki is an AI Assurance Architect and Product Engineering Lead based in Abu Dhabi. He designs how agentic AI systems are evaluated, red-teamed, governed and kept reliable, and engineers the AI products behind them, for enterprise and government programmes across the UAE and GCC.
- Long bio
- Satya Prakash Solanki is an AI Assurance Architect and Product Engineering Lead based in Abu Dhabi, with more than 13 years in quality engineering and AI product architecture across healthcare, finance, education, and oil and gas. He leads AI assurance and governance across an enterprise AI product portfolio: evaluation pipelines and release gates for LLM and agentic systems, AI red-teaming mapped to OWASP LLM Top 10, NIST AI RMF and MITRE ATLAS, guardrails, agent reliability and observability. He also builds multi-agent systems with LangGraph, CrewAI and MCP, runs an application security testing practice, and supports technical bids for enterprise and government AI assurance programmes. He was named Best Employee of the Year in 2024 and 2025.
Speaking
Planning an event or a leadership briefing?