Kiran Yenaganti
Founder · AI Governance ResearcherMy work focuses on how agents act, when they need human approval, and how to verify their results. I explore these questions through DataGovernAI and the HOMININI prototype.
Testing how AI workflows account for the work they report.
A report said more than the evidence supported.
While developing HOMININI, I found that its verifier accepted added claims such as “all tests passed,” even though the workflow had not established them. I repaired that path and added targeted regression tests.
The report could include a success statement without the evidence to justify it.
Add negative tests for the specific claims that had been accepted.
Local regression tests support this repair. Broader verification remains an open problem.
Policy Enforcement, Monitoring, and Risk-Based Human Escalation
An action can be permitted and still rely on incomplete or outdated information. My manuscript examines how evidence checks and human escalation can guide an agent at execution time.
Explore AI runtime governance research →Read on ResearchGateApplication no. 19/269,037 · Patent application, not a granted patent
The application describes an architecture for coordinating specialized agents with real-time data integration, governance checks, audit records, and fallback mechanisms. It grew out of the virtual leasing use case that first drew me to this problem.
System and Method for Governed Multi-Agent AI Workflow with Real-Time Data Integration and Session-Aware Voice Automation.
Named inventor: Kiran Kumar Yenaganti
Nonprovisional utility application: 19/269,037
Filing date: July 15, 2025
In January 2019, I left India at 22. It was my first flight.
I wanted to build something of my own. That ambition took shape in Techworkify, a technology staffing business through HyperLedger OÜ in Estonia.
COVID and travel restrictions interrupted those plans. I later studied Entrepreneurship & Innovation at Hult in San Francisco and founded DataGovernAI LLC in 2025.
My interest in AI governance grew from a leasing-assistant concept. Looking up availability and scheduling a tour sounded straightforward. But what if the data was old, or an agent needed approval before making a commitment?
Those practical questions led to my patent application and research into multi-agent systems. HOMININI is where I test parts of that thinking, alongside my doctoral studies and independent research.
Doctoral studies in progress · AI emphasis
San Francisco
My technical background includes data pipelines, transformation and data quality. That experience informs my research into the information AI agents rely on before taking action.
Working on agent verification, orchestration or runtime governance? I welcome technical questions and research conversations.
Connect on LinkedIn