🔍 Actively seeking LLM Engineering roles in the UK (Post-Study Work Visa ’26)
🔭 Current Focus:
Working on Model Context Protocol (MCP).
I’m actively looking for roles in LLM Engineering—fine-tuning, RAG systems, scalable serving—with eligibility to work in the UK through 2026.
👯 Open to Collaborate On:
– Transformer fine-tuning pipelines
– Hugging Face tokenization & model‐optimization
– Productionizing LLM APIs
– Retrieval-Augmented Generation (RAG), LangChain, Vector Database
– Implementing Multi Modal Chat with Gradio
– Evaluating frontier & open-source models
– Evaluvating models for specific project using Scale, Hugging Face and Vellum leaderboards
– Generating Meeting Minutes using LLM Audio file
– Creating audience-specific brochures via scraping & prompting
📫 Connect:
✉️ korarishi@gmail.com • LinkedIn • GitHub
| Modeling & NLP | **Machine Learning | Languages & Tools |
|---|---|---|
| • Transformers (BERT, GPT) | • Supervised ML | • Python, JavaScript |
| • Prompt Engineering | • UnSupervised ML | • PyTorch, Vector Database |
| • Tokenizers & Pipelines | • PEFT (LoRA, QLoRA) | • Hugging Face Transformers |
| • RAG & Retrieval | • privacy-preserving AI | • Scikit-Learn, Gradio |
-
College Brochure using HuggingFace
Generate AI-powered summaries for college brochures using Hugging Face models to help students know about the courses in college instead of browsing through the website. -
Pipelines-using-HuggingFace
Hands-on guide to Hugging Face pipelines: loading pretrained models, tokenization, and GPU-accelerated inference. -
Tokenizers using HuggingFace
Deep dive into tokenizer internals (LLaMA, PHI-3, StarCoder2) with examples for encoding, decoding, and token visualization. -
Fashion Customer Service Assistant using HuggingFace
Domain-specific assistant demo: fine-tuned transformer serving as a customer-service chatbot. -
Meeting Minutes using recording from the meeting
Generate professional meeting minutes from audio recordings using state-of-the-art models from Hugging Face Transformers..
-
predictive model for medical insurance costs
This project aims to develop a predictive model for medical insurance costs using machine learning techniques, specifically linear regression. The model will leverage various features such as age, BMI, smoking habits, and geographical region to estimate insurance charges... -
Predictive Maintenance
This project focuses on developing a predictive maintenance model to estimate the Remaining Useful Life (RUL) of machinery components. By leveraging machine learning techniques, the model aims to predict when a component is likely to fail, enabling proactive maintenance and reducing downtime... -
Netflix-Data-Analysis
This project aims to analyze and visualize data from a Netflix dataset...
(More notebooks in my repo showcase end-to-end analytics & ML pipelines.) :contentReference[oaicite:4]{index=4}
- MSc Data Science, University of Essex (2024)
- AWS AI Practioner
- UK Post-Study Work Visa valid through 2026
- Pronouns: He/His
“Predictive analysis saves time, money, and lives.”