ANA — Enterprise Conversational AI & Control Assurance

About the Product

ANA (Adaptive Neuro Intelligent Assistant) is an enterprise intelligence and AI operating layer for control assurance in regulated industries. Publicly featured by Anaptyss, ANA automates complex risk and compliance assessments, acting as an intelligent partner for risk professionals.

My Role & Contributions

As the Lead AI Engineer, I contributed to the architecture and development of ANA, focusing heavily on multi-agent workflows, conversational AI, and enterprise automation. My work involved gathering requirements directly from business stakeholders and translating them into robust, mathematically sound agentic architectures.

Technical Challenges Solved

• Designing a cohesive system where over 14 distinct LLM agents collaborate and cross-verify each other to ensure high-fidelity control testing.
• Implementing complex tool calling, allowing agents to fetch, query, and validate real-world evidence autonomously.
• Optimizing latency and orchestrating parallel workflows to ensure real-time responsiveness in enterprise environments.
• Integrating Text-to-Speech (TTS) and Speech-to-Text (STT) models to enable seamless conversational interactions.

Technologies Used

Lessons Learned

Building ANA reinforced the necessity of structured evaluation and mathematically rigorous design when deploying LLMs in high-stakes environments. Simply wrapping an API is insufficient; creating a self-correcting, multi-agent architecture is essential for enterprise reliability.