Mitansh Panchal
Software Engineer
Summary
Backend-focused Software Engineer with 2+ years of experience designing and building scalable server-side systems and production AI integrations (Gemini, Claude), MCP protocol, RAG pipelines, and vector search. Track record of shipping resilient APIs and automation — from prescription scanning with 83% accuracy to autonomous finance agents with 28 MCP tools. Strong in Node.js, Python, and API design with a focus on performance, reliability, and system-level automation.
Skills
Projects
Node.js, TypeScript, autocannon, @modelcontextprotocol/server, zod
API performance analyzer MCP server with 7 tools for benchmarking, load, stress, spike, soak, smoke, and A/B testing
Python, MCP, SQLAlchemy, Pydantic, Alembic, PostgreSQL
Personal finance MCP server for AI agents and LLMs - transactions, budgets, loans, reports over stdio.
TypeScript
Open-source, self-hosted authentication microservice for your applications.
Next.js, React, TypeScript, Tailwind CSS, Zod, Recharts
AI-powered Q&A to learn about Mitansh's professional journey, skills, projects, and experience.
Next.js, React, TypeScript, Drizzle ORM, Tailwind CSS, Recharts, Framer Motion
Workout tracker built around progressive overload - routines, exercise logging, and progress analytics.
Next.js, React, TypeScript, Prisma, PostgreSQL, Neon, Tailwind CSS, Google GenAI, Resend, Twilio, Recharts, Framer Motion, Zod
Next-gen fitness organizational backend system powering internal processing.
Experience
- Engineered an AI-powered prescription scanning feature using Google Gemini API, achieving 83% extraction accuracy and reducing manual data entry by 60%.
- Developed an automated age verification system leveraging Gemini LLM models integrated with cron job capabilities, achieving a 12% improvement in process efficiency.
- Architected and implemented modular RESTful APIs for a new e-commerce ecosystem, supporting thousands of concurrent users.
- Optimized backend order processing workflows and implemented automated PDF generation services, improving system efficiency by 40%.
- Developed analytical reporting modules with real-time dashboards for accounting and operations teams, reducing manual report generation by 15+ hours per week.
- Built an AI-generated image detection system integrating third-party vision APIs, flagging non-compliant product images with 90%+ accuracy across the catalog.
- Engineered automated prompt evaluation pipelines that scored Gemini model outputs against defined benchmarks, improving response quality iteration speed by 3x.
- Authored 50+ SEO-optimized technical articles on solar panel and residential battery systems, increasing organic website traffic by 35% over 6 months.
- Managed lead distribution workflows and built automated sales performance reports in Google Sheets, reducing manual tracking effort by 10+ hours per week.
Education
The Maharaja Sayajirao University Of Baroda
Diploma In Computer Engineering
Grade: 69%
Activities
- Spearheaded technical knowledge-sharing sessions, providing cross-team training on newly architected features and internal system workflows.
- Class Representative (Diploma Computer Engineering) for 2 consecutive years.
- Conducted 1-hour Automated Technical Seminar on Linux & Open Source Development at MSU Baroda.
- Freelance technical consultant for ExpositARVR focusing on website performance optimization.
- Curated technical content for a Linux-focused YouTube channel with +40k views and 300+ watch hours.
Languages
English (Professional), Hindi (Fluent), Gujarati (Native)