- Portfolio: 0rajnishk.github.io
I bridge the gap between complex AI research and production-grade applications. My background spans corporate R&D at FedEx, specialized GenAI development at IIT Madras, and delivering over 6 end-to-end commercial applications as a technical consultant.
- R&D Intern at FedEx Smart Lab: Architected Sarathi AI, a fleet management ecosystem. Engineered a computer vision model for driver behavior monitoring, improving detection accuracy from 60% to 100% and integrating real-time IoT (Raspberry Pi) data.
- GenAI & Full-Stack Developer at SGBC (Brain Centre): Developed Neuro Voyager, a RAG-based research platform presented at NVIDIA GTC. Implemented NeuroGuardrails and Knowledge Graphs for complex query handling.
- ML Engineer at Desi Crew: Building Multilingual Voice-to-Voice HR Chatbots (supporting 10+ languages) and multi-agent web scraping pipelines for large-scale data extraction.
- End-to-End Delivery: Successfully delivered 6 Webapps and Native Mobile Applications for global clients.
- Specialization: High-impact MVP development, AI/ML integration into legacy systems, and scalable microservices architecture.
- Generative AI: RAG Pipelines (FAISS/Pinecone), NL2SQL Systems, Multi-Agent Orchestration (uAgents, LangChain), Prompt Engineering, NeMo Guardrails.
- Machine Learning: Computer Vision (Driver Pose Detection), Predictive Analytics, TensorFlow, PyTorch, Scikit-learn.
- Full-Stack Web: Microservices Architecture, FastAPI, Django, Vue.js, Docker, RESTful APIs, Google Cloud Platform (GCP).
- Mobile & IoT: Flutter (Dart), Real-time Socket connections, IoT Hardware Integration (Raspberry Pi/ECU data).
- Studify: Developed a centralized educational platform serving 10,000+ students with personalized learning resources.
- Sarathi AI: Real-time fleet monitoring app integrating vehicle diagnostics (ECU) and driver safety alerts. Achieved 100% accuracy in head pose detection.
- AgentFlow: Multi-Agent Supply Chain Intelligence system. Runner-up at Fetch.ai Innovation Lab Hackathon.
- Neuro Voyager: RAG-based Q&A system for brain research papers. Showcased at NVIDIA GTC.
Open for Contracts & Collaboration: [email protected]
Show some love by starring a few repos!


