Divyansh Gautam

AI Engineer Intern @ DEEPQ AI

About

I am an AI/ML enthusiast and current B.Tech CSE (AIML) student at IIIT Nagpur, passionate about building impactful solutions using Artificial Intelligence, Machine Learning, and Data Science. I have hands-on experience in Python, NLP, RAG-based architectures, Computer Vision, and Data Analytics, with projects spanning multiple domains: - RAG-based AI Chatbot with Streamlit UI and Mistral/FLAN-T5 backend - Real-time Player Re-identification using YOLOv8/v11 + ByteTrack/DeepSORT - Speech Emotion Detection using Deep Learning - Image Forgery Detection using CNN architectures (ResNet, VGG16, MobileNet) with ELA - Plant Disease Classification, Customer Churn Prediction, and Lead Scoring models I enjoy solving challenging problems, experimenting with new AI techniques, and creating practical applications that bridge research and real-world impact. Core Skills:Python, Pandas, NumPy, Scikit-learn, TensorFlow, PyTorch, Streamlit, OpenAI API, FAISS, YOLOv8/v11, DeepSORT, ByteTrack, RAG, NLP, Data Analytics, SQL, Git

Country

India

City

Kota

Industry

Computer Software

Skill

Machine Learning, Deep Learning, AI Agents, Generative AI, Natural Language Processing (NLP), Computer Vision, AIML, Microsoft Azure, Document Outsourcing, Synthetic Data Generation, Data Pipelines, User Interaction, Problem Solving, Telemetry, Generative Adversarial Networks (GANs), Chatbots, Conversational AI, Artificial Intelligence (AI), Large Language Models (LLM), Retrieval-Augmented Generation (RAG)

Experience

AI Engineer Intern

DEEPQ AI

2025-7 - 2026-5 · 11 mos

• Developed Document AI systems (NLP + OCR-based IDP pipelines) using OCR (Surya), LLM pipelines, and Microsoft Azure Document Intelligence, enabling automated extraction of structured data from unstructured financial and KYC documents reducing end-to-end document processing time by 30%. • Built ML models (5+ use-cases: Customer Attrition, Predictive Triggers, Next Best Offer) using advanced feature engineering (RFM, temporal, ratio features), improving overall model performance by 15%. • Architected scalable end-to-end AI/ML pipelines (data generation → feature engineering → modeling → dashboards) for Machine Learning and LLM-based systems across BFSI and enterprise use cases. • Engineered multi-agent AI architectures using DSPy, CrewAI, and Claude-style orchestration, separating reasoning, execution, and control layers to enable scalable and modular AI workflows. • Designed a synthetic data generation framework using Ollama LLaMA3, creating domain-aware datasets with controlled correlations, improving robustness by 20%. • Deployed real-time AI dashboards on Replit with live demos and agent-based Q&A, translating model outputs into business insights and increasing stakeholder engagement by 25%. • Implemented agent telemetry systems (logging, approval tracking) for monitoring and debugging workflows.