Sudip Dhara

AI/ML Engineer with 2+ years of experience across professional and internship roles, developing AI-powered applications spanning Generative AI, Computer Vision, Deep Learning, NLP, and conversational AI. Experienced in building LLM-powered applications, RAG pipelines, multimodal AI workflows, image-generation solutions, computer vision systems, and conversational applications. Skilled in Python, PyTorch, TensorFlow, FastAPI, LangChain, and LangGraph, with hands-on experience integrating AI models into application-oriented solutions.

AI/ML Engineer - Onliest LLC - Visakhapatnam, India

(2026-05)

• Developing an AI-powered fashion design and recommendation platform using Large Language Models (LLMs), multimodal AI, computer vision, and image-generation models to create personalized saree and blouse designs.
• Building stateful conversational AI workflows using LangGraph, FastAPI, PostgreSQL, and Redis for contextual interactions, design recommendations, and real-time streaming responses.
• Implementing multimodal AI workflows for image understanding, retrieval-augmented generation (RAG), selective image editing, and regional fashion personalization.
• Developing AI-driven recommendation and conversational workflows to incorporate user preferences and contextual information into personalized design experiences.

Deep Learning Engineer - WizPhys AI Pvt. Ltd. - Pune, India

(2025-08 - 2026-05)

• Developed a computer vision-based physiotherapy assistance system using MediaPipe pose estimation and classification models to track and analyze patient movements in real time.
• Implemented movement recognition workflows for a neuro-gamified rehabilitation application, translating detected patient movements into corresponding in-game actions.
• Integrated AI models with smart-glove hardware to support real-time interaction and movement data processing for rehabilitation applications.
• Contributed to an AI-powered conversational chatbot using LangChain and LangGraph for contextual user assistance.

AI/ML Developer - Notional Systems Pvt. Ltd. - Kolkata, India

(2025-02 - 2025-08)

• Developed deep learning-based pose transfer and image translation workflows for human figure manipulation and background replacement.
• Built a pipeline to generate 3D shirt models from 2D texture pattern images for digital garment modeling.
• Worked on automating fabric texture extraction from images to streamline the 3D garment modeling workflow.

AI/ML Developer Intern - Pravaah Consulting - Bengaluru, India

(2024-06 - 2025-01)

• Developed YOLO-based computer vision models for facial disease identification using a custom medical image dataset and integrated a retrieval-augmented generation (RAG) workflow using LangChain and Pinecone.
• Developed PetGPT, a domain-specific conversational application by fine-tuning Llama 2 8B on a custom canine dataset for health, nutrition, and behavioral queries.
• Developed an X-ray fracture detection workflow using YOLOv8 and integrated Google Generative AI for contextual responses based on detected findings.

Deep Learning Intern - EI Networks Pvt. Ltd. - Delhi, India

(2024-05 - 2024-06)

• Trained YOLOv8 models on a custom dataset to detect driver behaviors including mobile phone usage, smoking, and headphone usage while driving.
• Built a vehicle identification pipeline combining number plate detection, OCR-based recognition, and image-based vehicle classification.
• Explored Variational Autoencoders (VAEs) for anomaly detection in electric vehicle battery performance data.

Data Scientist Intern - Pinkmoon Apps - Bengaluru, India

(2024-04 - 2024-04)

• Developed an automatic mantra counting system using Automatic Speech Recognition (ASR) to convert audio to text and track repetition counts, with Text-to-Speech (TTS) support.

NLP Intern - SkilldAI - Bengaluru, India

(2024-02 - 2024-03)

• Developed an OCR pipeline for handwritten text extraction from scanned documents using the Suriya library.

B.Tech - Computer Science and Engineering – Specialization in AI & ML - The Neotia University (2025)