Software engineer focused on building reliable, production-grade systems at the intersection of AI/ML and full-stack product engineering. I care about clean architecture, measurable impact, and shipping software that works at scale.
- 🧠 AI/ML — deep learning, computer vision, NLP, and agentic LLM systems.
- 🛠️ Full Stack — TypeScript/React/Next.js on the front, Python/FastAPI/Node on the back.
- 📊 Data & MLOps — end-to-end pipelines from data collection to deployed inference.
- 🧩 Product mindset — I build for the user, instrument for outcomes, iterate fast.
- 🧮 Strong CS fundamentals — active DSA practitioner (LeetCode, Striver, GFG).
🤝 Open To: Software Engineering · AI/ML Engineering · Full Stack roles · Open-source collaboration
Languages
Frontend
Backend & Databases
AI / ML
Cloud, DevOps & Tooling
| Domain | Proficiency | Details |
|---|---|---|
| Computer Vision | ⭐⭐⭐⭐⭐ | Gesture recognition, multilingual OCR, image-based detection pipelines with OpenCV |
| Deep Learning | ⭐⭐⭐⭐ | Multimodal deepfake & fake-news detection, CNN/transformer architectures |
| NLP | ⭐⭐⭐⭐ | Text classification, hybrid semantic search engines, content analysis |
| Agentic AI / LLMs | ⭐⭐⭐⭐ | LangChain-based research assistant, tool-using autonomous agents |
| Classical ML | ⭐⭐⭐⭐⭐ | 22+ end-to-end ML projects across regression, classification & clustering |
| MLOps & Deployment | ⭐⭐⭐⭐ | Model serving via FastAPI, Streamlit & Heroku; reproducible pipelines |
🧠 AI Career Platform
A full-stack, AI-powered platform that helps users navigate career growth — resume analysis, interview practice, and personalized guidance.
| Attribute | Detail |
|---|---|
| Stack | Next.js · TypeScript · Node.js · LLM APIs |
| Scale | Multi-module monorepo, modular services |
| Performance | Server-rendered UI, streamed AI responses |
| Security | Auth-guarded routes, environment-isolated secrets |
| Impact | End-to-end career tooling in one product surface |
| Repository | ai-career-platform-v2 |
Built as a production-oriented product with a clean separation between UI, service layer, and AI orchestration — designed to be extended feature-by-feature.
🕵️ Multimodal Fake News & Deepfake Detection System
A detection system that combines text and visual signals to flag manipulated media and misinformation.
| Attribute | Detail |
|---|---|
| Stack | Python · PyTorch · OpenCV · Transformers |
| Scale | Multimodal fusion across text + image inputs |
| Performance | Deep-learning classifiers tuned for detection accuracy |
| Security | Focused on trust & safety / integrity use-cases |
| Impact | Practical defense against synthetic media & fake content |
| Repository | Multimodals-Fake-News-Deepfake-Detection-System |
Explores multimodal learning by fusing linguistic and visual cues, demonstrating an applied research approach to a high-impact real-world problem.
🩺 Multilingual Medical OCR System
An OCR pipeline that extracts and understands text from medical documents across multiple languages.
| Attribute | Detail |
|---|---|
| Stack | Python · OpenCV · OCR engines · NLP |
| Scale | Multilingual document processing |
| Performance | Robust text extraction across varied inputs |
| Security | Handles sensitive medical text with care |
| Impact | Improves accessibility of medical records |
| Repository | MultiLingual_medical_OCR_System |
Combines computer vision and language processing to turn unstructured medical scans into structured, usable data.
🤖 Agentic AI Research Assistant
An autonomous research assistant that plans, searches, and synthesizes information using LLM tool-calling.
| Attribute | Detail |
|---|---|
| Stack | Python · LangChain · LLM APIs |
| Scale | Multi-step agent workflows |
| Performance | Tool-augmented reasoning & retrieval |
| Security | Controlled tool access & scoped execution |
| Impact | Automates literature and information gathering |
| Repository | Agentic-AI-Research-Assistant |
A hands-on exploration of agentic AI — chaining reasoning, memory, and external tools into a self-directed research loop.
🔐 Phishing Detection Web App
A web application that classifies URLs as legitimate or phishing in real time.
| Attribute | Detail |
|---|---|
| Stack | Python · scikit-learn · Flask/Streamlit |
| Scale | Real-time single-URL inference |
| Performance | Lightweight ML model for fast classification |
| Security | Directly targets a cyber-security threat vector |
| Impact | Helps users avoid malicious links |
| Repository | Phising_detection_webapp |
Demonstrates the full ML lifecycle — feature engineering, model training, and deployment into a usable web tool.
✍️ Gesture-Based Air Drawing
A computer-vision app that lets users draw in the air using hand gestures tracked via webcam.
| Attribute | Detail |
|---|---|
| Stack | Python · OpenCV · MediaPipe |
| Scale | Real-time video-stream processing |
| Performance | Low-latency hand tracking & gesture mapping |
| Security | Fully local, no data leaves the device |
| Impact | Touchless, accessible human-computer interaction |
| Repository | Gesture_Based_air_drawing_WebApp |
A fun, practical demonstration of real-time computer vision translating physical gestures into on-screen interaction.
MMM YYYY – MMM YYYY
Built and shipped features across the stack, contributing to AI/ML and product engineering initiatives.
- Developed and deployed application features from design through production.
- Collaborated with the team on data pipelines, model integration, and testing.
- Improved reliability and performance of core services.
Python TypeScript Machine Learning Full Stack Git
| Recognition | Details |
|---|---|
| 🤝 Pair Extraordinaire | Earned for co-authored commits & collaborative development on GitHub |
| ⚡ YOLO | GitHub achievement for merging without review — shipping with confidence |
| 🧩 DSA Practitioner | Solved structured problem sets across LeetCode, Striver & GeeksforGeeks |
| 🚀 60+ Repositories | Consistent building across AI/ML, full-stack & systems projects |
Amazon Web Services
Oracle
NPTEL
Cisco
Rahul_Kumar:
learning:
- Advanced LLM & agentic system design
- Scalable MLOps & model deployment
- System design & distributed systems
building:
- AI-powered full-stack products
- Computer-vision & detection pipelines
- Clean, production-ready web applications
exploring:
- Multimodal deep learning
- Retrieval-augmented generation (RAG)
- Open-source contribution
open_to:
- Software Engineering roles
- AI / ML Engineering roles
- Full Stack Development
- Collaboration on impactful projects
"First, solve the problem. Then, write the code."