name: Aadarsh Pandey
location: Bengaluru, India
role: Associate Software Engineer @ Accenture
education: B.Tech (CS) — Ajay Kumar Garg Engineering College, Ghaziabad
currently_building:
- AroundU — Hyper-local freelancing platform
- CodeRunr — Sandboxed code execution engine
- EcoMind AI — Research paper RAG system
currently_learning:
- Golang & Distributed Systems
- LangChain & LangGraph Agentic Workflows
- System Design at Scale
open_to_collaborate:
- Backend Systems (Spring Boot, Go)
- AI/ML Agents & RAG Pipelines
- Full-stack Mobile (Flutter)
fun_fact: "I'm an ambivert who writes rate limiters on weekends for fun.""Backend engineer focused on building fast, reliable systems at scale. I ship production-grade APIs, event pipelines, and mobile integrations that handle real-world traffic with strong reliability."
|
Associate Software Engineer • Oct 2024 – Present
|
|
SDE Intern • Mar 2024 – May 2024
|
|
SDE Intern • Sep 2023 – Mar 2024
|
📂 Detailed Project Breakdown (click to expand)
Full-stack platform connecting local service providers with clients in real-time
| Layer | Stack |
|---|---|
| Backend | Java 21 · Spring Boot 3 · JPA/Hibernate · PostgreSQL · Redis |
| Frontend | Next.js 14 · TypeScript · Tailwind CSS |
| Mobile | Flutter 3 · Riverpod · Dio · Google Maps · Geolocator |
Architecture Highlights:
- Job lifecycle state machine —
CREATED → OPEN_FOR_BIDS → BID_SELECTED → HANDSHAKE → IN_PROGRESS → PAYMENT → COMPLETED - OTP-secured handshake flow with 6-digit start/release codes, attempt locking & expiry
- Escrow payment system with lock/release mechanics
- Real-time bidding system with worker eligibility guards (on-duty, not blocked, no active job)
- Domain modules: Auth, Jobs, Bids, Chat, Payments, Reviews, Notifications, Location/Geocoding
- Monitoring: Prometheus + Micrometer metrics, Nginx reverse proxy
Judge0-style API for compiling & running code in isolated Linux sandboxes
| Layer | Stack |
|---|---|
| Backend | FastAPI · Python 3.12 · PostgreSQL 16 (async SQLAlchemy 2.0) · Celery + Redis |
| Sandbox | isolate (cgroup-based Linux isolation) · Docker |
| Mobile | Flutter 3 · Riverpod · Dio · Syntax Highlighting · Material 3 |
Highlights:
- 8 languages supported — C, C++, Python, JavaScript, TypeScript, Go, Rust, Java
- Async task queue with Celery workers for non-blocking execution
- Configurable resource limits — CPU time, wall time, memory, stack, file size, process count
- Token-based submission tracking with real-time polling
- AI code analysis panel (scaffold ready) for LLM-powered code review
- Full Swagger API docs with health checks
Production-ready backend simulating real-world healthcare architecture
| Component | Tech |
|---|---|
| Services | Auth Service · Clynic Service · Billing Service · Analytics Service · API Gateway |
| Stack | Java 21 · Spring Boot 3.x · Spring Security · JWT · gRPC · Kafka · PostgreSQL |
| Infra | Docker Compose · LocalStack (AWS sim) · GitHub Actions CI/CD · Shell IaC |
Highlights:
- gRPC inter-service communication (Clynic → Billing)
- Kafka event pipelines for async analytics processing
- JWT route-level access control with role-based authorization
- LocalStack infrastructure for ECS, VPC, RDS, MKS provisioning
- Full test coverage — unit, integration, embedded Kafka, MockMVC, gRPC mocks
- 3 production releases with CI-enforced coverage thresholds
AI-powered research assistant with multi-agent retrieval and vector search
| Component | Tech |
|---|---|
| Backend | FastAPI · LangChain · LangGraph · OpenAI · ChromaDB |
| Database | PostgreSQL (async) · Redis |
Highlights:
- Multi-agent RAG pipeline with retrieval, generation, and moderation agents
- ChromaDB vector store for semantic paper search
- Async FastAPI with SQLAlchemy 2.0 for paper metadata storage
Chat with any YouTube video using semantic search
Stack: Python · LangChain · Google Gemini · FAISS · YouTube Transcript API
- Extract transcripts → chunk → embed → semantic FAISS search → conversational Q&A
- Multi-query retrieval for better context understanding
Multi-agent system generating well-cited content on any topic
Stack: Python · LangGraph · Google Gemini · Tavily · Pydantic
- Two-stage architecture: First Responder → Tool Execution → Revisor → Final Answer
- Auto-generates search queries, adds numbered references with URLs
- Built-in reflection & critique system for iterative quality improvement
Browser automation for LinkedIn tasks using AI agents
Stack: Python · LangGraph · Playwright · Gemini · SQLite
- LangGraph agent with tool-call support for posting, messaging, connecting
- Playwright browser automation with DOM exploration
- Task logging with CSV/SQLite persistence
Native Android app with Kotlin, Jetpack Compose & Hilt
Stack: Kotlin · Jetpack Compose · Hilt DI · WorkManager · Android SDK 36
User management service with Spring Boot
Stack: Java · Spring Boot · JPA · PostgreSQL
Custom Redis server/client implementation from first principles
Stack: C++ · TCP Sockets · Systems Programming
- Built server & client from raw sockets with custom protocol
- First-principles approach to networking, error handling, and connection management
Personal portfolio with AI-powered virtual assistant
| Layer | Stack |
|---|---|
| Frontend | Next.js · TypeScript · Tailwind CSS |
| AI Backend | FastAPI · Gemini · NumPy RAG (no vector DB) |
- "Virtual Aadarsh" RAG assistant answers questions about my profile, projects & skills
- In-memory NumPy cosine similarity retrieval, strict persona prompt with refusal fallback
- Deployed on Railway (512MB budget)
Production-quality implementations of system design components
Stack: Java
- Rate Limiters — Token Bucket, Leaky Bucket, Fixed Window, Sliding Window
- Consistent Hashing — with virtual nodes for distributed systems
I write deep dives on backend engineering, system design, and distributed systems on Medium.
| Article | Topic |
|---|---|
| Building Rate Limiters from First Principles | Fixed Window, Sliding Window, Token Bucket, Leaky Bucket |
| Building Consistent Hashing from First Principles | Distributed Systems, Rebalancing, Virtual Nodes |
| Authentication & Authorization: The Security Backbone | Auth Mechanisms, Access Control |
| Understanding System Design | Scalability, Reliability, Large Systems |
| How Amazon's Notify Me Uses Observer Pattern | Design Patterns, Event-Driven |
| Concert Ticketing Systems at Scale | Load Balancing, Fairness, Failure Handling |
| Routing & Serialization | Controllers, Handlers, Payload Formats |
| Understanding HTTP for Backend Engineers | HTTP Methods, Headers, Statelessness |
💬 Open to impactful opportunities across backend systems, distributed architecture, AI engineering & product engineering.
📫 aadarshpandey.pandey@gmail.com
