Adithya Ramasubramaniam

Used Ollama and prompt engineering to standardize part data by checking dimensions and parameters, then generated normalized output with combined 2D/3D view, timestamp, and audit trail. Analyzed welding and geometry attributes, designed RAG system to recommend machines aligned to required standards, and built cost engine for part cost calculation. Developed review workflow code with human validation, structured JSON output pipeline from initial step through final step using orchestrator, and identified automation improvements for future AI integration. Built Streamlit-based ML advisor for specific datasets by organizing machine learning algorithms in one module and Ollama-driven model selection logic in separate module. Created agent-based 2D-to-3D image generation workflow using OpenAI, Python, and FreeCAD for instruction-driven 3D conversion. Defined agent roles for parameter extraction, LLM instruction generation, 3D execution, and consistency testing across 2D and 3D outputs. Built AI interview bot logic, partnered with lead on Streamlit interface design, deployed attendee through GitHub and Docker, and reviewed application logs from logs.txt.

• I did a drug price prediction project as my small internship project.I collected the data using snowflake,trained the data using linear regression.After training the data,I analysed the data using matplotlib library in python.This was a nice experience for me and I honed my machine learning skills further with the help of this internship.
• Core Engineering
• Key Skills: Python, Machine learning, Data analysis