Explore Full Systems Archive
Traverse all 17 autonomous systems, sub-200ms voice orchestrations, computer vision pipelines, and production case studies with category filtering and search.
AI engineer who builds agents that make decisions, with around two years across freelance client delivery and internships. Ships agentic systems end to end: a three-stage LLM guardrail agent packaged as an MCP plugin and adopted inside Claude Desktop, Cursor and Codex, multilingual voice agents, and full-stack AI products for 10+ clients. Comfortable owning a feature from prompt and tool design through deployment and evaluation. Looking for an early-stage team where I can build AI-native product.
Milestones, production deployments, and consulting record (2021 – 2026).
Began formal engineering degree, focusing on core programming in C and Python, data structures, algorithms, and Linux systems.
Built practical web utilities and automation scripts, deepening hands-on work with JavaScript, Python, REST APIs, and relational databases.
At Synar Technology, built production React.js and Flask web applications, improving render performance by 20% and earning promotion to Associate Software Engineer. Built mobile-first cross-platform apps at Kareng Technologies.
At IOCL, trained a linear regression forecasting pipeline on market data and constructed an automated Flask dashboard that reduced processing time by 40%. Concurrently deployed a 60 FPS YOLOv8 vision system.
Completed B.Tech capstone project with an automated ECG signal processing suite in Python, NumPy, and SciPy for QRS detection. Graduated in July 2025 and launched independent AI consultancy delivering production client platforms.
Architected Sentinel, an open-source 3-stage LLM safety guardrail shipped as an MCP plugin. Built multilingual autonomous voice agents on Vapi and Twilio, and engineered evaluation harnesses with fail-safes.

3-stage LLM security agent running as an MCP plugin in Claude Desktop, Cursor & Codex

Multilingual voice agent (English, Hindi & Telugu) with mid-call intent detection

Vision-guided DOM agent executing multi-step browser workflows
Frameworks, model protocols, and infrastructure proven across live client deployments.
Intercepts and blocks destructive AI agent terminal & file operations with zero unauthorized escapes.
Everyautonomoussystemrequiresadeterministicwitness: verify before trust, guarantee atomic rollback.
I engineer autonomous AI agents, conversational voice systems, and reliable full-stack applications. Focused on clean architecture, sub-second responsiveness, and software that delivers real-world value.
Open for full-time engineering roles, autonomous AI architecture, and high-impact product engineering. Connect directly through the verified pathways below.