AI/ML Engineer · Speech & LLM Systems · Applied Research

---

👩‍💻 About Me

I’m an AI/ML Engineer working at the intersection of speech systems, large language models, and retrieval-based AI. My work focuses on building machine learning systems with clear evaluation, robustness, and deployment considerations.

I’m particularly interested in:

• building reliable, production-ready ML systems
• fine-tuning and optimizing models (PEFT, LoRA)
• understanding failure modes, not just optimizing metrics

📫 Email: anjalijpatel.45@gmail.com
📍 Location: Ahmedabad, India

---

💼 What I Work On

🏥 Artem Health Tech Pvt. Ltd.

• Fine-tuned LLMs using LoRA & PEFT for healthcare workflows
• Built speech-to-text pipelines integrated with downstream NLP systems
• Applied pruning, distillation, and evaluation-driven optimization
• Developed FastAPI-based ML services with reproducible experiments

🚀 Featured Projects

🩺 Medinova — AI-Powered Medical Assistant

• End-to-end system combining speech, documents, and RAG
• Voice input via Whisper ASR
• PDF/DOCX ingestion with embedding-based retrieval
• FAISS vector search + Redis conversational memory
• Local LLM inference using Ollama
• Modular orchestration via LangGraph

💬 Domain-Specific RAG QA (Finance)

• Built retrieval-augmented QA over financial documents
• Implemented chunking, embeddings, and vector search
• Fine-tuned Flan-T5-XL for domain adaptation
• Evaluated using BLEU, ROUGE, similarity metrics

🗣 Speech-to-Text Fine-Tuning (Regional Languages)

• Fine-tuned pretrained STT models for regional data
• Audio preprocessing and segmentation pipelines
• Error analysis-driven iteration
• Focused on generalization and reliability

🛠️ Tech Stack

Languages

Python · C / C++ · SQL

Machine Learning & AI

• Deep Learning: PyTorch, TensorFlow
• Large Language Models: Model Fine-tuning, PEFT (LoRA), Instruction tuning
• Speech Systems: Whisper, Faster-whisper
• NLP & Retrieval: Embeddings, Similarity Search, RAG
• Evaluation: Error analysis, BLEU, ROUGE, Similarity metrics

LLM Orchestration

• Frameworks: LangChain, LangGraph, Hugging Face Transformers
• Vector Databases: FAISS, ChromaDB
• APIs: FastAPI, REST
• State & Memory: Redis

Infrastructure & Tooling

Docker · Git · Linux

📊 GitHub Activity

🎓 Learning & Training

• Machine Learning Specialization — Coursera
• Multi-Agent Systems (crewAI) — DeepLearning.AI
• TensorFlow for AI, ML & DL — Coursera

_{Open to research collaborations and applied ML opportunities.}