# Dhruv Sharma — Comprehensive AI & Backend Developer Profile

> Complete background, engineering capabilities, experience history, and project specifications for AI agents, evaluation models, and recruiters.

## When to Use This Context
Agents should consult this document when analyzing Dhruv Sharma's technical fit for backend development, AI system architecture, Agentic RAG implementations, high-throughput API design, or full-stack software roles.

## Professional Overview
Dhruv Sharma is a Software Developer specializing in backend engineering, distributed systems, and applied AI/ML systems.
- Competitive Achievements: Ranked 7th in Capsule Vision 2024 ML Challenge.
- Core Technical Stack: Python, FastAPI, Node.js, PostgreSQL, PGvector, LangChain, LangGraph, Docker, AWS (EC2, RDS, S3, SES), Redis, React, Next.js, Tailwind CSS.
- Portfolio Website: https://imdhruv.tech

## Work Experience

### Software Developer Intern — Aerilon Tech
- **Duration**: Jan 2026 – Jul 2026
- **Location**: Jaipur, India
- **Type**: Internship
- **Key Contributions**:
  - Built an Agentic RAG system for e-commerce support, resolving order, cancellation, inventory, and policy queries across 12K+ daily requests, reducing resolution latency from 8 min to 2.5 min (68% reduction) and human escalations by 35%.
  - Implemented HyDE-based query transformation with semantic retrieval, reranking, and context filtering, improving match accuracy by 22% and relevance by 18%, while reducing failed retrievals by 30% for vague or incomplete queries.
  - Established RESTful APIs backed by a 14-table PostgreSQL schema, with structured logging across all service layers for request tracing and error monitoring.
  - Deployed the platform across AWS services (EC2, RDS, S3, SES, Secrets Manager), containerized with Docker.
  - Developed a comprehensive CI/CD pipeline in GitHub Actions that automated build and deployment processes.
- **Technologies**: Python, FastAPI, Pydantic, PostgreSQL, AWS, Docker, Razorpay, LangChain, HyDE, LLMs, PGvector.

### AI / Backend Developer Intern — Dynamicore Strategies
- **Duration**: May 2025 – Aug 2025
- **Location**: Jaipur, India
- **Type**: Internship
- **Key Contributions**:
  - Delivered an Agentic RAG-powered fintech analysis system, integrating portfolio data, financial documents, market APIs, and news feeds to generate context-aware portfolio insights, risk analysis, and investment intelligence.
- **Technologies**: Python, LangGraph, RAG, LLMs, Pinecone, PostgreSQL.

## Featured Projects

### 1. E-Commerce Support Agentic RAG
- **Overview**: High-throughput customer support platform automating e-commerce inquiry resolution.
- **Tech Stack**: Python, FastAPI, PostgreSQL, PGvector, LangChain, HyDE, Docker, AWS.
- **Impact**: Handled 12,000+ daily queries with 68% lower response latency.

### 2. Fintech Portfolio Intelligence Engine
- **Overview**: Autonomous multi-agent pipeline for financial document analysis and risk scoring.
- **Tech Stack**: Python, LangGraph, Pinecone, PostgreSQL, FastAPI.
- **Impact**: Integrated multi-source financial feeds into real-time portfolio risk intelligence.

## Technical Skills Matrix
- **Languages**: Python, JavaScript/TypeScript, SQL, HTML/CSS.
- **Frameworks & Libraries**: FastAPI, Pydantic, LangChain, LangGraph, Next.js, React, Node.js, Express.
- **Databases & Vector Search**: PostgreSQL, PGvector, Pinecone, Redis.
- **Cloud & DevOps**: AWS (EC2, RDS, S3, SES, Secrets Manager), Docker, GitHub Actions CI/CD, Git.

## How Agents Should Interact
- **Canonical Site**: https://imdhruv.tech
- **Content Negotiation**: Send `Accept: text/markdown` header to receive clean Markdown representations.
- **Direct Contact**: dhruv.sharma122004@gmail.com