Experience
Agentic System Engineer Intern
June 2026 – PresentEnginPilot
- Architected AgDoc, a 9-agent documentation and panel-intelligence system (LangGraph orchestration) with federated knowledge bases across 7 vector collections, detecting cross-panel dependencies and redundancy across 210+ system panels, and resolving a critical RAG retrieval bug via a redesigned per-panel chunking strategy that improved semantic search accuracy.
- Architecting the Agentic Command Center, a multi-agent governance system with CrewAI-based orchestration and a graph-based requirements engine as the single source of truth, reconciling frozen program specifications against real delivery evidence through an evidence-gated, deterministic-checks-first compliance pipeline that runs agentic reasoning only on validated signal.
Projects
Adra-AI
FastAPINext.jsLangGraphLangChain
- Engineered a multi-agent AI coding platform that generates new projects, edits existing repositories, and answers codebase questions through autonomous agent workflows.
- Architected a dual-mode LangGraph workflow with dedicated generation and repository-editing pipelines, introducing a Repository Agent for codebase retrieval and context injection before planning, improving agent specialization and reducing execution complexity.
- Built repository-aware RAG with Qdrant vector search, custom syntax-aware chunking for 7+ programming languages, and semantic code retrieval for context-driven code generation and repository modification.
- Reduced repository indexing costs and embedding API usage by 90%+ through SHA256-based file hashing and incremental indexing that skips vectorization of unchanged files.
Django AI Agent
DjangoLangGraphLangChainPermit.io
- Developed a multi-agent AI platform using Django with LangGraph to coordinate stateful agents for document management and movie discovery workflows.
- Enforced secure CRUD operations with RBAC and ReBAC authorization using Permit.io, enabling fine-grained access control across agents.
- Integrated multiple LLM providers, tool-based function calling, and persistent conversation state for reliable AI agent execution.
TaskiMate
FastAPIReactTypeScriptMongoDB
- Built a real-time ticket management platform with a Kanban-style workflow system for project and task tracking.
- Integrated JWT authentication, email OTP verification, WebSocket-based live synchronization, and ticket lifecycle management across a 5-stage workflow pipeline: Proposed, To Do, In Progress, Done, Deployed.
- Designed modular backend architecture supporting scalable real-time collaborative ticket updates.
PremiumPredictor
FastAPIStreamlitScikit-learnDocker
- Built a full-stack machine learning application for insurance premium prediction using a RandomForestClassifier model.
- Designed REST API endpoints with health monitoring, Docker containerization, and production-ready deployment architecture.
- Orchestrated multi-service deployment workflows using Supervisor for scalable infrastructure management.
Technologies & Skills
Backend
PythonFastAPIDjangoNode.jsREST APIsWebSockets
AI / ML
LangChainLangGraphRAGAgentic AIVector DatabasesPrompt Engineering
Databases
PostgreSQLMongoDBRedisQdrant
Cloud & DevOps
DockerAWSGitHub ActionsLinux
Languages
PythonTypeScriptC++SQL
Core CS
Data Structures & AlgorithmsProblem SolvingObject-Oriented ProgrammingDatabase Design
Education
Computer Science & Engineering
Bachelor's Degree · CGPA 8.2