Full-Stack & Systems Engineer | AI / ML & Distributed Workflows | Cloud Architecture
Final-Year B.Tech Computer Science & Engineering @ KIIT University, Bhubaneswar
I design and build resilient, deterministic full-stack applications, distributed AI data pipelines, and blockchain systems. My engineering philosophy centers on anti-hallucination code execution, strongly typed full-stack architectures (TypeScript + Next.js/React + Prisma), and high-throughput Python/Go backends deployed across cloud-native environments (Docker, Kubernetes, AWS, Vercel, Render).
- π» Core Domains: Distributed Systems, Full-Stack Web Development, Data Engineering, LLM Agentic Workflows, Smart Contracts.
- π οΈ Current Focus: Architecting enterprise-grade event-driven microservices, AST-sandboxed AI code generation, and financial econometrics.
- π― Engineering Principle: Always prefer deterministic computation and verifiable code execution over raw LLM guessing.
| Domain | Technologies & Frameworks |
|---|---|
| Languages | TypeScript, JavaScript, Python 3.12, Java, Solidity, SQL (PostgreSQL, MySQL), HTML5/CSS3 |
| Frontend | React 18, Next.js (App Router), Tailwind CSS, Material UI, Monaco Editor, Electron |
| Backend & APIs | Node.js, Express, FastAPI, Uvicorn, Gunicorn, Celery, RESTful APIs, WebSockets, JWT RBAC |
| Databases & ORM | PostgreSQL, Prisma ORM, MongoDB (Mongoose), SQLAlchemy, SQLite, Redis |
| Cloud & DevOps | Docker, Kubernetes, AWS, Google Cloud Platform (GCP), GitHub Actions (CI/CD), Vercel, Render |
| AI / ML & Data | Google Gemini API, Groq LLM Streaming, Scikit-Learn, EGARCH Volatility Modeling, Pandas, NumPy, BeautifulSoup |
| Web3 & Blockchain | Solidity (^0.8.0), Web3.js, Ethers.js, OpenZeppelin, Truffle Suite, Hardhat |
| Project | Tech Stack | Highlights & Architecture | Live Demo |
|---|---|---|---|
| FlashAI | Next.js, TypeScript, Groq LLM, SuperMemo-2, Prisma, Tailwind | AI-powered spaced-repetition study engine with ultra-low latency Groq LLM streaming, SM-2 retention algorithm, and flashcard deck management. | Live App |
| csv-qa-agent | Python, Pandas, Groq API, AST Sandbox, Streamlit | Rooman AI Challenge Winner: Zero-hallucination conversational tabular data Q&A agent executing sandboxed Pandas queries with AST security guards. | GitHub Repo |
| mini-erp-crm | React 18, TypeScript, Express, Prisma, PostgreSQL, Tailwind | Full-stack wholesale/distribution ERP & CRM operations portal featuring inventory tracking, multi-line delivery challans, automated GST PDF invoices, and JWT RBAC. | Live Frontend Β· Backend API |
| investor-sentiment-streamlit | Python 3.12, Streamlit, EGARCH, Scikit-Learn, YFinance | Indian stock market volatility analyzer (Nifty 50, Sensex) pairing asymmetric EGARCH(1,1) econometric modeling with Random Forest directional forecasting. | Live App |
| cursor-ide-clone | Electron, React, TypeScript, Monaco Editor, Python runtime | AI-powered desktop code editor for Python with real-time Monaco editing, inline ghost-text code completions, interactive selection rewrites, and terminal execution. | GitHub Repo |
| deal-proposal-automation | Python, MCP Tool Calling, HTML5/CSS3, Telegram Webhook | Zero-delay B2B proposal and SOW generator converting raw post-sales call discovery notes into executive proposals & 1-tap mobile approval cards in < 3 minutes. | GitHub Repo |
| automation_hub | FastAPI, Docker, Kubernetes, Celery, Redis, Gemini AI | Production-grade distributed web scraper & PDF intelligence engine with Google Gemini AI document summarization and async task queues. | GitHub Repo |
| drug-supply-chain | Solidity, Web3.js, Ethers.js, React 18, Node.js, MUI | Pharmaceutical counterfeit prevention and IoT cold-chain temperature compliance DApp with on-chain threshold breach alerts and role-based access control. | GitHub Repo |
| GIT-1ST | Java 22, Python 3.10+, DSA Patterns, Unit Testing | Curated reference repository for Data Structures & Algorithms patterns (Sliding Window, Kadane's, DP, Trees) with automated test suites. | GitHub Repo |
- Generative AI Solutions β Google Cloud Virtual Internship (fine-tuned models, optimized token throughput, prompt engineering pipelines).
- Process Mining & Workflow Analytics β Celonis Virtual Internship (diagnosed enterprise bottleneck latency, improved throughput by 15%).
- Open Source Engineering β Active contributor maintaining production-tested repositories across full-stack, AI, and blockchain ecosystems.
- π LinkedIn: linkedin.com/in/rahul-roy-362a12256
- π§ Personal Email: rahulroy2259@gmail.com
- π» GitHub Profile: @Rahul03ll
- π Location: Bhubaneswar, Odisha, India
