Generative AI Systems • Safety & Governance • Privacy
Apex, North Carolina
Venkata Rami Reddy Kallu
Senior AI Engineer | Generative AI Systems, Safety & Privacy
I design and build production-grade generative AI systems with
enforceable safety, privacy, and governance guarantees —
spanning LLM tool execution, runtime validation,
and adversarial robustness across text and audio.
My work sits at the boundary between engineering and applied research,
where production constraints require runtime enforcement rather than best-effort heuristics.
Areas I work on:
Policy-gated tool execution for LLM agents
Deterministic validation of tool inputs and outputs
Evidence-backed summarization (anti-hallucination by construction)
Privacy-preserving detection of synthetic voice
I’m especially interested in systems that are auditable, reproducible,
and resistant to misuse even when models behave unexpectedly or adversarially.
Now
What I’m actively shipping and writing:
PolicyGraph: a reproducible safety runtime for MCP tool execution
OWASP-aligned evaluation + contract tests for agent safety
Follow-up work on runtime enforcement patterns (Paper #2 direction)
Privacy-Preserving Synthetic Voice Detection (Filed systems patent)
I’m a co-inventor on a filed systems patent focused on
privacy-preserving synthetic voice detection — detecting AI-generated/spoofed speech while
minimizing speaker identity leakage.
Detect AI-generated or spoofed speech reliably
Preserve speaker anonymity and sensitive identity attributes
Support deployment in privacy-sensitive and regulated settings