Miloni Panchal

AI Researcher | AI & Data Science Graduate | Ex-Research Intern @ BARC | ML Stack • LLMs • Explainable AI | Turning Ideas into Impact

📧 [email protected]

---

About

GenAI researcher and ML engineer building intelligent systems that matter. From fine-tuning LLMs for storytelling to developing explainable AI for medical diagnosis, I create AI solutions that are both powerful and interpretable. Passionate about pushing the boundaries of what's possible with generative AI.

---

Experience

AI Research Intern - BARC | Oct 2024 - May 2025
• Engineered AI pipeline processing 25K+ variants from 1L+ samples
• Achieved 96.51% accuracy in pathogenicity prediction
• Deployed web UI with LIME/Grad-CAM explainability

Practical Training - BARC | July 2024 - Aug 2024
• Built wavelet-based CNN for anomaly detection (97.14% accuracy)
• Reduced false positives by 20% through advanced preprocessing

---

Key Projects

Energy Consumption Prediction | LSTM time-series model, RMSE: 172.5, reduced wastage by 12-15%

Customer Retention Analysis | Cohort analysis identifying 15-20% higher repeat rates

Document Routing System | Multi-agent AI with Mistral LLMs, reduced processing time by 75%

LLM Storyteller | Fine-tuned GPT-2 + real-time generation, 85%+ relevance

---

Tech Stack

Languages: Python, Java, C, JavaScript, SQL
AI/ML: TensorFlow, Keras, Transformers, Hugging Face, LangChain
Tools: Streamlit, Power BI, MongoDB, MATLAB
Specialized: Multi-Agent AI, LIME, Grad-CAM, BorutaPy

---

Achievements

🥇 MSME Hackathon 3.0 - Technical Round Qualifier
🎖️ BARC Research Commendation - AI pathogenic variant detection
🏅 Multiple Awards - Intercollegiate events

---

Open to: AI/ML roles, research collaborations, hackathons

"Turning data into insights, insights into impact."