Patents & IP

8 non-provisional USPTO filings powering deterministic AI platforms.

This portfolio represents a unified line of work: deterministic, offline, and privacy-preserving AI systems. Each patent is tied to a real product surface—with a working or planned demo and a research paper targeting journal or conference submission. Application numbers are listed for transparency. Full claims and specifications are shared privately with hiring teams and collaborators.

Title
Application No.
Area & Status
Edge-Deployed LLM V2: Privacy-Preserving Offline Architecture with Self-Forgetting Memory and On-Device Alignment Debugger
19/268,142
Edge AI · Privacy · Alignment
Filed 2025 · Pending examination
PromptPilot: Deterministic Prompt Evaluation & Orchestration Engine
19/269,169
LLM Safety · Prompt Governance
Filed 2025 · Pending examination
AI Risk Navigator: Deterministic LLM Risk Scoring & Triage
19/275,864
AI Safety · Governance · Compliance
Filed 2025 · Pending examination
LLM Code Safety Auditor: Offline, Rule-Based Source Code Evaluation and Remediation Engine
19/283,236
AppSec · Static Analysis
Filed 2025 · Docketed · Ready for examination
AutoRedact AI: Deterministic Sensitive-Data Redaction for Unstructured Content
19/281,903
Privacy · Data Protection
Filed 2025 · Pending examination
TraceSafe AI: Audit-First LLM Tracing & Policy Enforcement Layer
19/281,714
Observability · Governance
Filed 2025 · Pending examination
Self-Healing Prompt Engine (SHPE)
19/281,647
Reliability · Prompt Recovery
Filed 2025 · Pending examination
AutoJudge: Offline, Rule-Based LLM Output Adjudication Engine
19/279,355
Evaluation · Safety · Governance
Filed 2025 · Pending examination

How this IP maps to product strategy

Together, these patents form an opinionated product platform around AI safety, Edge AI, and governance: offline edge inference, deterministic safety engines, privacy-preserving telemetry, prompt governance, code safety, and lineage. They’re not isolated ideas—they’re building blocks for an enterprise AI platform roadmap.

In interviews or partner discussions, I map these filings directly to your stack: where an Edge AI runtime fits, how safety engines bolt onto existing observability, and how governance layers give legal and security the guarantees they need.