Building LLM, RAG, and production ML systems that hold up at scale. 5 years across research and industry.
About
Senior AI Engineer with 5 years of experience taking LLM, RAG, and ML systems from prototype to production. Currently architecting intelligent document processing pipelines at Cognisure AI — reducing manual extraction time by 70% and semantic retrieval latency by 60%.
Research and industry in parallel: as AI Researcher & Assistant Professor at Jain Deemed University, I led 10+ applied AI/ML projects, published peer-reviewed work in IEEE and Springer, and hold a published patent in AI-based traffic enforcement. Simultaneously delivered 90+ AI solutions to international clients.
From LoRA fine-tuning to agentic multi-agent systems, from Qdrant vector search to Azure OpenAI deployments — I build systems that hold up in production at scale.
Experience
Cognisure AI
May 2025 – Present
Full-time
Gen Hypothesis
Feb 2024 – Apr 2025
Consulting
Jain Deemed University
Dec 2021 – May 2025
Research & Academia
Fiverr
Dec 2020 – Dec 2021
Freelance
Einfach BA
Dec 2020 – Apr 2021
Full-time
Skills
AI & Machine Learning
LLM Engineering · RAG Systems · Computer Vision · Agentic AI · LoRA Fine-tuning · Prompt Engineering · NLP
Frameworks & Libraries
LangChain · LlamaIndex · PyTorch · TensorFlow · Scikit-learn · OpenCV · YOLO · MLflow · CrewAI · LangGraph
Infrastructure & Platforms
FastAPI · Docker · Azure OpenAI · AWS · Qdrant · Pinecone · FAISS · n8n · MongoDB
Languages
Python · SQL · JavaScript · Java · MATLAB
Projects
FastAPI service comparing CVs against job descriptions using LLM reasoning (OpenAI / Gemini) with a deterministic fallback scorer. Modular parsing, extraction, scoring, and recommendation services — Dockerized with Pydantic schemas and rotating logs.
Real-time computer-vision motion-detection server on FastAPI — supports file upload, live camera capture, and URL streaming with async processing. Frame-difference algorithm with configurable sensitivity, SQLAlchemy metadata storage, and background retention cleanup.
NLP product that auto-tags daily journal entries with multi-dimensional sentiment analysis, shipped at Gen Hypothesis.
Master's dissertation — computer vision and pattern recognition on 150+ handwritten samples in Python and MATLAB, training an ML model to predict Big Five personality traits.
Research
Peer-reviewed work across IEEE, Springer Nature, and CRC Press, plus a published patent — spanning applied ML, computer vision, security, and IoT.
Conference
CML 2025 · Springer
Feb 2026
Journal
Int. J. System Assurance Eng. & Mgmt · Springer Nature
Jan 2025
Patent
App. No. 202441096943A · Published, awaiting grant
Dec 2024
Real-Time Law Enforcement and Road Safety with Smart Traffic Enforcement System: Integration of Artificial Intelligence and Vision Models
Conference
ICACRSE 2024 · IEEE
May 2024
Conference
PEEIC 2023 · IEEE
Dec 2023
Conference
PEEIC 2023 · IEEE
Dec 2023
Book Chapter
Chapman & Hall/CRC · ISBN 9781003319238
Jul 2023
Certifications
Amazon Web Services · Coursera · Jun 2025
Google Cloud Training · Aug 2024
AWS & DeepLearning.AI · Jun 2024
Google Cloud Training · Jun 2023
DeepLearning.AI & Stanford Online · Mar 2023
Cleared · Jun 2020
Contact
Open to senior AI engineering roles, consulting engagements, and research collaborations. Based in Bengaluru, India — available globally for remote work.
GitHub
github.com/DrakunalScholar
Google Scholar