Available for Full-Stack & GenAI Roles

Harsh Mann

Full-Stack AI Engineer

Full-Stack AI Engineer specializing in production RAG systems and MERN applications. At Design Avenue, I build AI agents that automate compliance work — policy generation, evidence review, and repository classification — integrated across AWS, GCP, GitHub, and Jira. I pair hybrid retrieval and vector search with clean API design and modern React frontends to ship context-grounded, scalable products.

Resume

LinkedIn GitHub Email

Technologies I work with

React.js

Next.js

TypeScript

JavaScript

Redux Toolkit

Tailwind CSS

Node.js

Express.js

REST APIs

Socket.io

JWT

RBAC

Java

MongoDB

$vectorSearch

React.js

Next.js

TypeScript

JavaScript

Redux Toolkit

Tailwind CSS

Node.js

Express.js

REST APIs

Socket.io

JWT

RBAC

Java

MongoDB

$vectorSearch

Aggregation Pipelines

MySQL

SQL

AWS

GCP

/Git/GitHub

RAG Systems

Hybrid Retrieval

Vector Search

BM25

Reranking

Embeddings

OpenAI

Claude

Gemini

Aggregation Pipelines

MySQL

SQL

AWS

GCP

/Git/GitHub

RAG Systems

Hybrid Retrieval

Vector Search

BM25

Reranking

Embeddings

OpenAI

Claude

Gemini

Professional Experience

Enterprise RAG pipelines, AI agents, and compliance automation at Design Avenue — plus the earlier roles that built my engineering foundations.

Full-time · Feb 2026 - Present
Software Developer — Design Avenue
- Develop full-stack modules for HRMS and GRC platforms — role-based access control, automated workflows, and compliance risk management features.
- Architect RAG-based AI pipelines — document ingestion, structural chunking, vector search, and metadata filtering — to deliver context-grounded LLM responses across product areas.
- Engineer AI agents for policy generation and compliance automation, integrating LLMs with external APIs and enterprise platforms (AWS, GCP, GitHub, Jira) for automated evidence review and repository classification.
- Optimize backend aggregation pipelines and API routes to cut latency across high-traffic enterprise modules.
React.jsNode.jsExpress.jsMongoDBRAG
Internship · Aug 2025 - Jan 2026
Software Developer Intern — Design Avenue
- Co-engineered core microservices and database schemas for ChemDoc and HRMS platforms using the MERN stack.
- Developed secure, high-performance REST APIs and reusable React.js components styled with Tailwind CSS.
- Built Retrieval-Augmented Generation (RAG) features utilizing vector search to enable automated, intelligent document parsing.
- Engineered automated workflows and robust role-based access protocols to handle multi-tenant enterprise data securely.
React.jsNode.jsMongoDBTailwind CSS

Projects

Full-stack systems and AI pipelines I've designed end to end — click any featured build to read the full case study: the problem, the architecture, and the takeaways.

ContextRAG

A production-grade hybrid RAG studio: dense vector search fused with BM25 keyword retrieval, cross-encoder reranking, and an enforced citation contract so every LLM claim is traceable to a source chunk.

Hybrid retrieval — 2048-dim Atlas Vector Search fused with BM25 via Reciprocal Rank Fusion
Cohere Rerank v4.0 Pro cross-encoder prunes irrelevant context before generation
Enforced citation contract with per-answer citation-match score and an interactive evidence drawer

React.jsTypeScriptExpress.jsMongoDBVector SearchBM25RerankingLLM APIs

GitHub (Source)

DevTinder

A full-stack developer networking platform: interest-based matching, connection requests, real-time Socket.io chat with presence tracking, and Razorpay-powered premium memberships.

Socket.io chat with online/offline presence and last-seen — socket-level auth validation, status kept off hot DB paths
Compound-indexed MongoDB: paginated feed via $nin/$ne and duplicate-proof connection requests via .pre middleware
Razorpay payments end to end — order creation, webhook handling, verification, and history

React.jsRedux ToolkitNode.jsExpress.jsMongoDBSocket.ioJWTRazorpay

GitHub (Source) Live Demo

AI Chat TOC

A browser extension adding an auto-generated, drag-and-drop Table of Contents sidebar to ChatGPT, Claude, Gemini, Perplexity, and Grok — used by 500+ people across three extension stores.

500+ active users across the Chrome, Edge, and Firefox add-on stores
Live TOC across ChatGPT, Gemini, Perplexity, Claude, and Grok — with search, answer previews, and markdown export
Near-zero-lag updates via 500ms throttled DOM observation with idle safety checks

JavaScriptWebExtensions APIHTMLCSS

GitHub (Source) Live Demo

Technical Expertise

Skills organized the way I architect systems — typed frontend state, backend services, data infrastructure, and retrieval pipelines.

01 / FrontendTyped, state-driven UIs

React.jsNext.jsTypeScriptJavaScriptRedux ToolkitTailwind CSS

02 / Backend & SystemsAPIs, realtime & auth

Node.jsExpress.jsREST APIsSocket.ioJWTRBACJava

03 / Databases & CloudData modeling & infrastructure

MongoDB$vectorSearchAggregation PipelinesMySQLSQLAWSGCP/Git/GitHub

04 / AI & RetrievalContext-grounded LLM pipelines

RAG SystemsHybrid RetrievalVector SearchBM25RerankingEmbeddingsOpenAIClaudeGemini

Get in Touch

Have a question or want to discuss a project? I'm always open to new opportunities and collaborations.