Ved.

AI / ML & Backend Developer

Backend Developer

Ved Pandya ยท Ahmedabad, India

About

Here is a little background

Hey ๐Ÿ‘‹ I'm a Backend & AI engineer based in Ahmedabad, India. I build scalable systems, agentic GenAI workflows, and cloud-native apps โ€” from media intelligence platforms to news backends serving 100K+ visitors. I love turning complex data problems into reliable, production-grade software.

Bachelor of Technology in Computer Engineering โ€” Charusat University (CGPA: 8.6)

Experience

Backend Developer

Adosphere Brandings Pvt. Ltd.

Jan 2023 โ€“ Present

  • Architected and built scalable, high-performance applications with a strong focus on user experience and reliability.
  • Led end-to-end project delivery, providing technical direction and driving product engineering decisions.
  • Designed and implemented advanced agentic workflows using state-of-the-art GenAI models, automating complex manual processes with near-human accuracy.
  • Led client communications and stakeholder management, bridging technical and non-technical teams.
  • Mentored junior developers, fostering technical growth and improving team productivity.

Skills

TypeScript
JavaScript
Python
Node.js
Next.js
React
FastAPI
LangChain
LangGraph
PostgreSQL
MongoDB
Redis
Docker
AWS
Qdrant
Neo4j
GraphQL
BullMQ

Projects

Project 1: Listening Tool

AI-Powered Media Intelligence & Insights Platform

Scalable pipeline ingesting multi-platform social data into an AI processing layer with transcription, sentiment, clustering, and semantic retrieval.

  • โ€“Multi-platform ingestion (YouTube, Instagram, Facebook, X) via scrapers and queues.
  • โ€“AI workflows for transcription gating, language detection, and multilingual transcription.
  • โ€“Embeddings in Qdrant, clustering, topic modeling, and rule-based regional reach scoring.
  • โ€“Guardrails, tracing, and LLM observability with LangChain and Langfuse.

Project 2: Docxrok

GenAI Document Intelligence Platform

OCR-enabled document pipeline with hybrid Agentic AI and GraphRAG for structured extraction and relationship-aware search.

  • โ€“Processes PDFs, images, DOCX, and text into structured data.
  • โ€“Hybrid Agentic AI + GraphRAG for chunking and contextual retrieval.
  • โ€“Neo4j integrated with vector embeddings for semantic search.
  • โ€“MCP-based custom search and automated DOCX report generation.

Project 3: Offbeat Stories

News Platform

Full-stack news backend serving 100K+ monthly visitors with ISR, Cloudflare caching, and high uptime.

  • โ€“Scalable, secure content delivery architecture on AWS.
  • โ€“ISR and Cloudflare caching for performance and SEO.
  • โ€“Collaborated with product and editorial teams on engagement features.

Project 4: Amdavad Na Garba

Event Ticketing System

QR-based ticketing and payment platform processing 200K+ tickets across large-scale Navratri events.

  • โ€“Online purchases and on-site validation with custom QR scanner app.
  • โ€“Secure transactions, real-time validation, and fraud prevention.
  • โ€“Optimized for high concurrent traffic during peak entry hours.

Project 5: Meta Ads Tool

Facebook Data Integration

Automated Meta Ads API integration with scheduled analytics sync to Google Sheets.

  • โ€“Fetches Facebook post data via Meta Ads API.
  • โ€“Google Sheets integration for link-driven scheduling.
  • โ€“Cron-based periodic updates reducing manual reporting.

Contact

I have got just what you need. Lets talk.