Arsh

Built a multi-agent evaluation engine with LangGraph, orchestrating parallel specialist agents for automated codebase assessment, resume verification, and real-time voice interviews using OpenAI Realtime API. Designed pre-interview intelligence pipelines that infer technical depth from resumes and generate targeted interview questions using structured LLM outputs, checkpointing, SQS-based async workflows.

Developed

Contextinator, a hybrid code search tool combining Tree-sitter AST chunking with embedding-based semantic search using ChromaDB and fast local grep, used by AI agents to navigate and evaluate unfamiliar repositories.

Lead Applied AI Engineer and Cofounder at UVAL.AI (2025-01 – Present)

• Built a multi-agent evaluation engine with LangGraph, orchestrating parallel specialist agents for automated codebase assessment, resume verification, and real-time voice interviews using OpenAI Realtime API.
• Designed pre-interview intelligence pipelines that infer technical depth from resumes and generate targeted interview questions using structured LLM outputs, checkpointing, SQS-based async workflows.
• Developed Contextinator, a hybrid code search tool combining Tree-sitter AST chunking with embedding-based semantic search using ChromaDB and fast local grep, used by AI agents to navigate and evaluate unfamiliar repositories.

Software Engineering Intern at Graphic Era University (2025-03 – 2025-08)

• Built an internal placement portal used by the university's placement department for managing drives, student applications, and company coordination using Next.js, Hono.js, and Prisma ORM.
• Implemented role-based auth (NextAuth), S3 document management, Nodemailer notifications, Excel report generation, and analytics dashboards with Recharts + TanStack Table for placement statistics and department-level metrics.

B.Tech in Computer Science and Engineering – Graphic Era Deemed to be University (2022 – 2026)