Senior AI Lead | Applied AI Scientist | GenAI | Computer Vision
I work as a senior AI lead and applied machine learning scientist. Over the last seven years, I have worked on Generative AI, RAG, LLM agents, computer vision, and healthcare intelligence, with a strong bias for systems that are useful outside a demo.
About
I like working where research and engineering meet. My current work includes chatbots, knowledge assistants, RAG pipelines, vector search, multimodal systems, and LLM agents.
I still enjoy being close to the models and architecture, but my role now also includes technical direction, delivery decisions, and helping teams build AI systems that behave reliably. Earlier work includes image enhancement, super-resolution, image harmonization, healthcare recommendation systems, government chatbots, and pandemic dashboards.
Experience
Leading work on chatbots, knowledge assistants, RAG systems, vector databases, multimodal LLM workflows, agent-based systems, reliability guardrails, and deployment planning.
Worked on applied ML problems for clients and products, compared features against strong research baselines, and helped turn promising ideas into usable product features.
Worked on computer vision models for image enhancement and super-resolution, deployed ML models as APIs, and built backend infrastructure for AI applications.
Built recommendation and decision-support systems, healthcare and government chatbots, and predictive models for medical and pandemic datasets.
Built AI-based investigation recommendation systems and chatbot solutions for nationwide healthcare management.
Skills
Generative AI, LLMs, RAG, LangChain, AI agents, prompt design, model evaluation, applied research, and ML workflows.
Image enhancement, super-resolution, image harmonization, quality metrics, object placement, lighting, and artifact analysis.
Vector databases, embedding pipelines, metadata-aware search, Django APIs, AI services, observability, deployment automation, and technical leadership.
Pandas, NumPy, Django, and backend services that help make machine learning models usable in products.
Education
Feb 2021 - Dec 2023
Artificial Intelligence
Feb 2018 - Aug 2018
Big Data Analytics
Aug 2013 - Jul 2017
Information Technology
Research & IP
Abhijat Chaturvedi and co-authors
Evolving Technologies for Computing, Communication and Smart World: Proceedings of ETCCS 2020
Abhijat Chaturvedi and co-authors
Abhijat Chaturvedi and co-authors
Abhijat Chaturvedi and co-authors
Abhijat Chaturvedi and co-authors
Projects
Developed an AI-driven system to automatically organize visual content into coherent storyboards, enabling rapid visualization of narratives from large image collections. Applied computer vision for scene analysis and automated sequencing to capture logical flow and thematic continuity across marketing, creative content, and digital storytelling workflows.
Worked on models that blend objects into generated backgrounds more naturally by fixing color, lighting, and texture mismatches.
Designed a multimodal framework to quantify perceptual image quality for AI-generated and natural images. Built a scoring system covering naturalness, object placement, lighting balance, edge sharpness, and realism using segmentation maps and object masks. Integrated vision-language models (CLIP, ViTs) for semantic alignment between text prompts and generated images, combined into a composite quality metric spanning statistical, perceptual, and aesthetic dimensions.
Developed a deep learning–based super-resolution model to enhance image quality and detail recovery from low-resolution inputs. Designed a transformer-based architecture with Flash Attention for faster inference and lower GPU memory usage. Benchmarked against SwinIR, SwinHAT, and RDN on DIV2K, Set5, and Set14, achieving competitive PSNR and SSIM with fewer parameters using combined perceptual and L1 loss.
Built a comprehensive and interactive dashboard tracking COVID-19 statistics — cases, active cases, recoveries, and deaths — across India and the world.
Built a medical decision-support system for case history analysis, medication suggestions, and treatment guidance.
Built a recommendation system to suggest similar medical investigations based on patient age, gender, speciality, medical history, and clinical conditions.
Developed a general English-language chatbot for nationwide blood search, enabling users to find blood donors and banks across India through natural language queries.
GitHub
These repositories are pulled from GitHub automatically, so the list stays current as public work is added.
Blog
I write literature on Medium as a hobby. It is a different side of me from the AI work, and this section updates as new pieces are published.
Contact
I am open to conversations around AI leadership, applied research, Generative AI, RAG, LLM agents, computer vision, ML platforms, and thoughtful AI product work.