Sachin Asokan
I am Sachin Asokan, a passionate Machine Learning Engineer and Data Scientist with a Master of Science in Industrial Engineering from Purdue University.
About
I am Sachin Asokan, a passionate Machine Learning Engineer and Data Scientist with a Master of Science in Industrial Engineering from Purdue University. I specialize in leveraging state-of-the-art machine learning models and cloud technologies to build cost-efficient, scalable AI solutions that drive business growth and improve customer experiences.
With hands-on experience in deploying vector databases, optimizing LLM inference, automating data pipelines, and generating insights through advanced computer vision and graph neural networks, I thrive on solving complex problems and delivering measurable results. I am eager to bring my skills in programming, cloud platforms, and data science to innovative projects and challenging environments.
- 2 years' experience
- Joined August 2025
Industry experience
- Software & Internet
- Healthcare
- Life Sciences
- Manufacturing
More experts in Columbus, United States
Work experience
Achieved 97% cost savings by replacing AWS Kendra with a self-hosted ChromaDB vector database and FastAPI application on a t3.medium EC2 instance. Reduced costs by 77% by deploying a quantized Llama2 model for CPU-based inference and implemented caching for faster response. Integrated chatbot functions including LLM API calls, ChromaDB document retrieval, and DynamoDB using custom Lambda orchestration. Built LangChain Agents to orchestrate chatbot functions, improving customer satisfaction score by 10%.
Developed a novel social media analytics tool using computer vision and LLMs to provide competitive insights boosting marketing and sales. Automated image dataset creation with custom scrapper and multithreading, halving processing time. Applied CLIP for zero-shot image validation, eliminating need for object detection models, saving five months in development time. Processed multimedia data with llava, Videollama, and Whisper to extract descriptive text, increasing sales by 10%. Built RAG system with Llama2 and LangChain for social media QA, raising customer reach by 15%. Created knowledge graphs from large text corpora using REBEL, enabling better data interpretation. Trained GraphSAGE GNN for link prediction with 84% accuracy, producing actionable business insights.
Collaborated with Purdue University and Bayer to develop a mixed linear model to predict crop yield using genetic markers and environmental features, achieving an R-squared value of 0.3.
Education
Hire Sachin Asokan today
Post a job and invite Sachin, or contact them directly. No hidden commission.






