Chirag Dhongadi
About
Detail
India
Contact Chirag regarding:
Flexible work
Starting at
USD1.21/hour
id_card
Internships
Starting at
USD193/month
Résumé
Jobs
verified_user
0% verified
- KData Science InternKokoro.DoctorJan 2026 - Current (6 months)Developing a LangGraph-based prescription generation system that processes multi-format patient reports, extracts critical clinical information, and generates structured, context-aware medical prescriptions. Designed multi-step LLM workflows using Lang Chain and LangGraph to analyze laboratory reports, medical history, and diagnostic data for accurate clinical summarization. Implemented Retrieval-Augmented Generation (RAG) pipelines with Hugging Face models to ensure evidence-grounded, reliable, and context-aware medical recommendations. Built and deployed secure, production-ready FastAPI endpoints to enable real-time prescription generation and scalable clinical decision support integration.
- AMachine Learning InternAICTE – TechSaksham (Microsoft & SAP CSR Initiative)Dec 2024 - Feb 2025 (3 months)Developed a deep learning-based plant disease detection system under the TechSaksham AI program. Trained a CNN using TensorFlow/Keras to classify potato leaves into Healthy, Early Blight, and Late Blight. Implemented data preprocessing and image augmentation to improve generalization and reduce overfitting. Achieved 97% classification accuracy and deployed the model in an interactive application.
Education
verified_user
0% verified
- GB.Tech in Computer Science and Engineering (Specialization: Data Science)Gyan Ganga Institute Of Technology And SciencesSep 2022 - Current (3 years 10 months)Bachelor of Technology (B.Tech) in Computer Science and Engineering (Data Science Specialization) Developing strong foundations in Machine Learning, Deep Learning, Artificial Intelligence, Data Structures, Algorithms, and Statistical Methods, with hands-on experience in building real-world AI and data-driven applications.
Projects (professional or personal)
verified_user
0% verified
- VVoxMed AIApr 2026 - Current (3 months)• Built an AI-powered multimodal medical assistant using voice and image inputs. • Integrated Whisper, Llama Vision, and Groq APIs for speech and image understanding. • Developed an end-to-end Speech → LLM → Voice interaction pipeline. • Enabled low-latency AI responses for real-time conversational diagnosis support.
- MMath Mentor AIMar 2026 - Current (4 months)• Built a multimodal AI tutor for solving mathematical problems from text, image, and voice inputs. • Integrated OCR and Speech-to-Text pipelines for handwritten and spoken query understanding. • Developed a LangGraph + LangChain workflow for step-by-step mathematical reasoning. • Implemented Human-in-the-Loop validation and session memory to improve response reliability.
- MMedRAGJan 2026 - Current (6 months)Built a Retrieval-Augmented Generation (RAG) system to generate structured medical revision notes from multi-format clinical documents. Implemented OCR with Tesseract and document ingestion for PDF and image preprocessing. Designed FAISS-based semantic retrieval using HuggingFace embeddings to ground LLM outputs. Integrated Ollama-hosted MedGemma with prompt engineering for adaptive revision generation. - Utilized cloud platforms to architect scalable infrastructure, ensuring high availability and efficient deployment of the RAG system across diverse environments. - Developed and maintained interactive dashboards that visualize model performance and retrieval accuracy, providing stakeholders with clear insights into system efficacy. - E
- CCalorie AIOct 2025 - Current (9 months)Built an XGBoost regression model to predict calories using age, gender, height, weight, duration, and heart rate. Performed preprocessing, feature engineering, and EDA using Pandas and Matplotlib to enhance model performance. Achieved R2 = 0.9988 and MAE = 1.48 through hyperparameter tuning and cross-validation. Deployed the trained model using Flask to deliver real-time calorie predictions via an interactive web interface.
- PPotato Leaf Disease DetectionFeb 2025 - Current (1 year 5 months)Built a CNN model using TensorFlow/Keras to detect Healthy, Early Blight, and Late Blight potato leaves. Applied data preprocessing and image augmentation techniques to improve model generalization and reduce overfitting. Achieved 97% classification accuracy and deployed the model through an interactive Streamlit web application.