JaiminPanchal
I'm anAI Engineer
I build production LLM systems — RAG pipelines, tool-calling agents and evaluation harnesses — on a foundation of 2.5+ years shipping backend software at scale.
// about
About me
II bbuuiilldd iinntteelllliiggeenntt ddiiggiittaall eexxppeerriieenncceess —— ffuullll--ssttaacckk ssyysstteemmss,, AAII--ppoowweerreedd pprroodduuccttss,, aanndd iinntteerrffaacceess tthhaatt mmoovvee.. WWiitthh aa ffooccuuss oonn eennggiinneeeerriinngg tthhaatt sshhiippss aanndd ddeessiiggnn tthhaatt ffeeeellss aalliivvee,, ii ppaarrttnneerr wwiitthh tteeaammss aanndd cclliieennttss wwhhoo wwaanntt ttoo ssttaanndd oouutt aanndd pprroovvee wwhhaatt mmooddeerrnn ssooffttwwaarree ccaann bbee.. LLeett’’ss bbuuiilldd ssoommeetthhiinngg iinnccrreeddiibbllee ttooggeetthheerr!!
Based in India · Remote.
2.5+
Years engineering
10K+
Orders/month at scale
5
Platforms shipped
6
Languages spoken
// engineering journey
How I got here
PHP / Laravel
Where I started shipping real backends
→Full-Stack
Owning features end to end
→React / Next.js
Modern, motion-rich frontends
→APIs & Integrations
Connecting systems reliably
→AI Integration
LLMs inside real products
→RAG & Vectors
Grounded, retrieval-based AI
→GenAI Applications
AI that does useful work
→AI Product Engineering
Where I’m focused now
// skills & stack
What I work with
The tools I reach for across the stack — from interfaces and APIs to databases and applied AI.
AI / LLM
Languages
Backend & Serving
Databases
Cloud & DevOps
Testing
// career
Experience
Sep 2025 — Present
AI / LLM & Backend Engineer
Freelance · Remote
Independent AI/LLM and full-stack engagements — production RAG pipelines, tool-calling agents and evaluation harnesses.
- Ship AI/LLM work with Python/FastAPI, LangChain/LangGraph, pgvector/ChromaDB and OpenAI/Gemini — production RAG pipelines, tool-calling agents and evaluation harnesses.
- Built the Laravel + MySQL backend and REST APIs for Spot It Now, a cross-platform wildlife-sighting app — geospatial feeds, photo uploads, OTP auth and push notifications for Android & iOS, plus a moderation admin panel.
Jul 2023 — Aug 2025
Software Engineer (Backend Focus)
Sapphire Software Solution · Ahmedabad, India
Delivered client platforms end-to-end across SaaS, e-commerce, fintech and social products.
- Delivered five client platforms end-to-end (Laravel, MySQL) — all five remain live in production.
- Moved order fulfilment into queued background workers for a Shopify backend sustaining 10,000+ orders/month; optimized MySQL via indexing and query restructuring.
- Engineered event-driven Laravel backends (queues, events, secure webhooks) powering live bid broadcasting, a transactional wallet ledger and automated revenue-split payouts.
- Employee of the Quarter (2025) · Emerging Star Award (2024).
// education & credentials
Foundations
2019 — 2023
B.Tech in Computer Science & Engineering
Indus University, Ahmedabad
CGPA 8.89 / 10
Employee of the Quarter
Sapphire Software Solution · 2025
Emerging Star Award
Sapphire Software Solution · 2024
Generative AI Mastermind
Outskill · 2026
IBM Project Manager — Professional Certificate
Coursera · 2026
// selected work
Projects
A selection of products I've designed, engineered and shipped.
DocIntel
Multimodal document-intelligence API — extracts schema-validated fields from PDFs and images with a vision LLM, per-field review flags and rules-based validation.
DocuMind AI
Production document-Q&A SaaS — FastAPI, pgvector RAG, JWT auth, Stripe billing, async Celery workers and Langfuse tracing.
ResolveAI
LangGraph support agent that resolves cases via tool calling and hybrid retrieval — pgvector (HNSW) fused with Postgres full-text via RRF, cross-encoder reranked, with deterministic policy guardrails.
ShopSense
E-commerce API with natural-language product search and an LLM shopping assistant — pgvector semantic search, function calling, Stripe checkout and Socket.io real-time updates.
FeedbackIQ
Multi-tenant Laravel SaaS API that ingests feedback and classifies sentiment, themes and urgency via an LLM dispatched to queued Redis workers — OpenAPI, Pest tests and Dockerized CI.
OrderMesh
Event-driven order-fulfilment platform using Spring Boot, Kafka and the Saga orchestration pattern.
// ai lab
The AI Lab
I don't just call ChatGPT. I build retrieval-augmented systems that ground answers in real data — embeddings, vector search and LLMs wired into products that need to be correct, not just confident.
RAG Pipelines
Hybrid search — pgvector fused with Postgres full-text via RRF, then cross-encoder reranked.
Tool-Calling Agents
LangGraph agents with deterministic policy guardrails enforced in code, not prompts.
Evaluation Harnesses
RAGAS & DeepEval — faithfulness, answer relevancy and context precision, wired into CI.
Vector Search
pgvector (HNSW cosine) and ChromaDB embeddings, benchmarked against dense-only baselines.
Structured Outputs
Schema-validated LLM outputs (Pydantic) with per-field review flags and fallbacks.
Multi-Model Serving
OpenAI, Google Gemini and local Ollama for cost-controlled, reliable inference.
retrieval-augmented generation · request flow
// what i build
The things I make
I build web products, AI-powered applications and practical developer tools — end to end.
AI-Powered Applications
LLM integrations, RAG systems, AI assistants and intelligent workflows that do real work — not demos.
Full-Stack Products
Complete web applications from frontend to backend, database, auth and deployment.
Developer Tools
Internal tools, dashboards and automation that remove friction from real workflows.
Career & Productivity Tools
Job tracking, resume analysis and AI-assisted career workflows — like the system powering this site.
Currently building
// let’s talk
Let’s build something
Have a product to ship, an AI idea to explore, or a role to fill? I’m open to select freelance work and full-time opportunities.