Founding AI Engineer at Armoriq, building runtime security and governance for LLM agents. Intent-based MCP security gateway serving 60,000+ policy checks a day across a 1,000+ tool registry, with tool-drift detection cut from 24 hours to under 10 minutes.
Shipping Voice_ and Humanizer_, native macOS apps on WhisperKit, CoreML, MLX, and Apple FoundationModels. Research across simulation-based inference, agentic AI evaluation, faithful generation, and multimodal medical imaging.
| Year | Paper | Venue | Focus |
|---|---|---|---|
| 2026 | Content Determination Is the Faithfulness Bottleneck | INLG 2026, Utrecht · oral | Faithfulness in data-to-text NLG, GlassGate framing |
| 2026 | Agentic Artificial Intelligence: Architectures, Taxonomies, and Evaluation of LLM Agents | arXiv:2601.12560 | Survey and evaluation framework for LLM agents · with R. Buyya |
| 2026 | Quantum Annealing Enhanced Reinforcement Learning for Accurate Remaining Useful Lifetime Prediction | arXiv:2606.18503 | Quantum-assisted RL for predictive maintenance |
| 2026 | RAID: Semantic Graph Diffusion for True Cold-Start and Cross-Lingual Forecasting | arXiv:2606.16925 | Graph diffusion, cold-start and cross-lingual forecasting |
| 2026 | MA-SBI: Misspecification-Aware Simulation-Based Inference via Side-Channel Guidance | arXiv:2606.16923 | Robust SBI under model misspecification |
| 2025 | When Algorithms Manage Humans: A Double Machine Learning Approach | arXiv:2512.22290 | Nonlinear effects of algorithmic control on gig worker performance and wellbeing |
| 2025 | A-QCF-Net: An Adaptive Quaternion Cross-Fusion Network | arXiv:2512.21760 | Multimodal liver tumor segmentation from unpaired CT/MRI · +7% Dice on LiTS and ATLAS |
Research assistant at NIT Tiruchirappalli and Georgia Tech. Full list on Google Scholar.
| Company | Role | What I Built |
|---|---|---|
| Armoriq | Founding AI Engineer | Intent-based MCP security gateway · 60k+ daily policy checks · 1,000+ tool registry · tool-drift detection from 24h to under 10 min |
| CompasIQ | Backend Engineer | FastAPI microservices on GCP · 50k daily compliance checks at sub-150ms p95 |
| Ionic Protocol | Engineer | LangGraph multi-agent RAG on AWS · 500 qpm · 4x TTFT improvement |
| Project | Stack | What It Does |
|---|---|---|
| Voice_ | Swift · WhisperKit · CoreML | Real-time on-device transcription for macOS |
| Humanizer_ | Swift · MLX · Apple FoundationModels | On-device AI-text detection and rewriting |
| AgentWeaver | Python · LangGraph | Multi-agent orchestration framework |
| aiprocesstransformer.com | Python · FastAPI · TypeScript | Nine-agent document pipeline, public beta |
Research PyTorch · MLX · CoreML · WhisperKit · Apple FoundationModels
Simulation-based inference · ABM · Diffusion · RAG · LangGraph
Apple Native Swift · SwiftUI · CoreML · MLX · Xcode
Backend Python · FastAPI · TypeScript · Node.js · Go · Docker
GCP · AWS · PostgreSQL · pgvector · Redis · SQLite
Open to research collaboration and engineering roles · github.com/akv2011