Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm an AI/ML Engineer with hands-on experience building production GenAI and ML systems across enterprise AI consulting in India, and I am currently an engineering intern in Dublin. I specialize in RAG pipelines, LoRA and PEFT, and multimodal AI, working across LangChain, HuggingFace Transformers, PyTorch, FAISS, Chroma, and Pinecone. I have delivered end-to-end ML workflows including model development, evaluation, and explainability via SHAP, plus deployment through Docker, GitHub Actions, and Azure. I am pursuing an MSc in AI at DCU. Notable projects include a RAG pipeline that reduced invoice processing errors by 30% and a retrieval system achieving 87% precision across 50,000+ records.…I'm an AI/ML Engineer with hands-on experience building production GenAI and ML systems across enterprise AI consulting in India, and I am currently an engineering intern in Dublin. I specialize in RAG pipelines, LoRA and PEFT, and multimodal AI, working across LangChain, HuggingFace Transformers, PyTorch, FAISS, Chroma, and Pinecone. I have delivered end-to-end ML workflows including model development, evaluation, and explainability via SHAP, plus deployment through Docker, GitHub Actions, and Azure. I am pursuing an MSc in AI at DCU. Notable projects include a RAG pipeline that reduced invoice processing errors by 30% and a retrieval system achieving 87% precision across 50,000+ records.

MAANASA PANCHAKARLA

AI Engineer, Data Scientist, Full Stack Developer, +2





I'm an AI/ML Engineer with hands-on experience building production GenAI and ML systems across enterprise AI consulting in India, and I am currently an engineering intern in Dublin. I specialize in RAG pipelines, LoRA and PEFT, and multimodal AI, working across LangChain, HuggingFace Transformers, PyTorch, FAISS, Chroma, and Pinecone. I have delivered end-to-end ML workflows including model development, evaluation, and explainability via SHAP, plus deployment through Docker, GitHub Actions, and Azure. I am pursuing an MSc in AI at DCU. Notable projects include a RAG pipeline that reduced invoice processing errors by 30% and a retrieval system achieving 87% precision across 50,000+ records.…I'm an AI/ML Engineer with hands-on experience building production GenAI and ML systems across enterprise AI consulting in India, and I am currently an engineering intern in Dublin. I specialize in RAG pipelines, LoRA and PEFT, and multimodal AI, working across LangChain, HuggingFace Transformers, PyTorch, FAISS, Chroma, and Pinecone. I have delivered end-to-end ML workflows including model development, evaluation, and explainability via SHAP, plus deployment through Docker, GitHub Actions, and Azure. I am pursuing an MSc in AI at DCU. Notable projects include a RAG pipeline that reduced invoice processing errors by 30% and a retrieval system achieving 87% precision across 50,000+ records.

Available to hire

I’m an AI/ML Engineer with hands-on experience building production GenAI and ML systems across enterprise AI consulting in India, and I am currently an engineering intern in Dublin. I specialize in RAG pipelines, LoRA and PEFT, and multimodal AI, working across LangChain, HuggingFace Transformers, PyTorch, FAISS, Chroma, and Pinecone.

I have delivered end-to-end ML workflows including model development, evaluation, and explainability via SHAP, plus deployment through Docker, GitHub Actions, and Azure. I am pursuing an MSc in AI at DCU. Notable projects include a RAG pipeline that reduced invoice processing errors by 30% and a retrieval system achieving 87% precision across 50,000+ records.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Language

English

Fluent

Work Experience

AI/ML Engineering Intern at Kefron Digital Limited

February 1, 2026 - Present

Contributed to building a RAG pipeline for invoice data extraction using LangChain, FAISS, Chroma, and Pinecone; reduced processing errors by ~30% vs OCR-only approach. Led NLP entity extraction models for vendor names, line items, and PO codes achieving 89% field-level extraction accuracy. Implemented an LLM monitoring layer with Azure Application Insights to track quality and hallucinations and supported containerisation/deployment via Docker and Azure Container Apps. Participated in Agile ceremonies and presented the RAG approach internally.

AI/ML Engineer at Soothsayer Analytics

January 1, 2024 - August 1, 2025

RAN LoRA and PEFT fine-tuning experiments on open-source LLMs using HuggingFace TRL; built a RAG-based document retrieval system for 50,000+ documents achieving ~87% retrieval precision. Developed a multimodal sentiment analysis system (text, audio, facial inputs) with SHAP explainability and deployed a demo via Gradio on HuggingFace Spaces. Implemented GDP trend forecasting with XGBoost and Random Forest across 15 countries. Established CI/CD for model versioning and automated retraining using GitHub Actions and AWS S3.