Back to search

Vishal T

Application Engineer — LLM & MCP Specialist

Bhopal, IndiaSouth AsiaMid-levelAvailable
Request an intro

Architects multi-provider AI systems solo — MCP servers to production, swappable by design.

  • Immediate to deploy
  • MCP server author
  • 7 production platforms — sole engineer
  • Multi-provider LLM architecture
  • Forward deployed engineer
  • RAG-powered recruiter chatbot
7
Projects on record
33
Tools and frameworks

Profile

AI-first application engineer building MCP servers and multi-provider LLM architectures across Gemini, OpenAI and Anthropic. Ships full-stack platforms on Next.js/TypeScript, NestJS and FastAPI, with real-time WebSocket pipelines and RAG search on PostgreSQL/pgvector. Owns delivery end to end — Docker, CI/CD, and production deployment on AWS and Azure.

Track record

  1. Provider-swappable by design

    Built an AI architecture where LLM and TTS providers switch on a single environment variable — no vendor lock-in anywhere in the stack.

  2. An ATS recruiters can talk to

    Shipped an AI-powered applicant tracking system with a RAG chatbot, letting recruiters query the candidate database in plain English.

  3. Decisions written down

    Authors MCP servers with explicit written rationale for every model-selection call — a decision record another engineer can pick up, not just working code.

  4. AI-native workflow, daily

    Works in Claude Code, Cursor and Windsurf for context engineering, and treats the tooling as part of delivery speed rather than as a novelty.

What stands out

Forward deployed engineer

Builds and deploys client-specific solutions — a RAG-powered recruiting ATS, legal-tech, financial services — each owned end to end for the engagement.

Multi-provider orchestration

AI pipelines that run across Gemini, OpenAI and Anthropic, with the provider as configuration rather than as architecture.

Full-stack execution

Owns frontend, backend and database design on every project, with zero hand-offs from architecture to deployment.

Careful system design

Versioned REST APIs, granular RBAC across roles, and automated SLA monitoring — engineering that anticipates edge cases before they become incidents.

Selected work

Echo Suite & Echo Being — AI meeting intelligence

Multi-product suite, sole engineer

  • Real-time in-meeting intelligence — suggestions, Q&A, Mermaid diagrams — over a Gemini Live WebSocket.
  • High-fidelity post-meeting transcription pipeline built on PydanticAI.
  • Fully provider-swappable across LLM and TTS vendors via one environment variable.

AI-powered applicant tracking system

RAG-powered recruiter chatbot

  • Modern ATS whose chatbot lets recruiters query the candidate database in plain English.
  • Granular RBAC across four roles, automated SLA monitoring, end-to-end recruitment flow tracking.

Titan Hub — document intelligence

  • Enterprise RAG platform supporting PDF, PPTX, DOCX and XLSX.
  • Routes image-rich documents to Gemini Vision and text-heavy ones to Docling, with chain-of-thought reasoning and source citations.
  • Real-time progress over Server-Sent Events.

Legal tech platform

  • Full-stack MERN platform simplifying tax, legal and financial services for SMEs.
  • Owned delivery from system design to production — schema, admin dashboard, REST API, n8n automation pipelines.

Financial services dashboard

Solo-built, deployed end to end

  • React and Redux Toolkit frontend with JWT and Passport.js auth over HTTP-only cookies.
  • Optimised MongoDB query layers.

Supply chain & inventory system

  • Full-stack inventory and billing solution with AI-based OCR on Gemini to digitise invoices and automate data entry.
  • Real-time stock transfer queue across factory, inventory and outlet modules.

E-commerce platform

Adopted by 10+ shops and online brands

  • Multi-tenant storefront with a companion React Native mobile app.
  • Full order flow — cart, checkout, payments, discounts, real-time admin notifications.

Certifications

  • Udemy Web Development Bootcamp (MERN)
  • Cisco Python Essentials 1 & 2
  • NeetCode DSA

Vishal is available through Alphanext.

Tell us the shape of the work — stack, timezone, how long — and we will set up a conversation. No account, no shortlist fee.

Request an introSearch the bench again