Data Scientist 4 YoE, experienced in building & deploying ML, DL, & GenAI models, with strong statistical analysis
India
#OpenToWorkI am a data scientist with 4 years of experience, currently working at Flutter International since November 2024. My career began after graduating in 2022, and I have since built expertise in AI, machine learning, and generative AI models across multiple companies in India. My work spans Python programming with Pandas, NumPy, and Scikit-Learn, as well as deep learning frameworks like TensorFlow and PyTorch. I have engineered solutions involving LLMs such as GPT and BERT, built chatbots, anomaly detection systems, and recommendation engines, and deployed models using AWS Sagemaker, FastAPI, and cloud platforms including AWS and GCP.
Data Science Tech or Product Management opportunity
Flutter International
• Engineered a smart knowledge assistant leveraging multi-model RAG architectures and vector database (FAISS) for semantic retrieval, enabling contextual reasoning across heterogeneous data sources • Built a time anomaly detection system for poker hands count prediction leveraging the N-BEATS model, improving detection of irregular play patterns and enhancing system monitoring via alerting method
Cars24
• Developed an LLM-driven chatbot with multilayer intent segmentation using transformer embeddings (BERT, MP- Net). Integrated CrewAI for agentic orchestration and real-time conversational control. • Fine-tuned all-mpnet-base-v2 for high-precision classification of domain queries. Implemented preprocessing, training, and evaluation pipelines with Scikit-learn and PyTorch. • Built a car-service history summariser using PDF-scraping, bigram-based NLP filtering, and LLM layers to produce structured summaries. Leveraged vector embeddings and retrieval for accuracy.
Paytm
• Built LLM-powered chatbot that seamlessly integrates OpenAI and Azure to deliver real-time responses, while simultaneously performing intent recognition and maintaining conversation context through LangChain Model • Designed an end-to-end XTTS model, leveraging GPU to minimize latency. Additionally, develop a FastAPI for deployment on EC2 instance. Leveraged AWS Sagemaker for GPU instance deployment for real-time application.
IIT Ropar TIF (AWaDH)
• Analyzed 20+ patent search for pre-existing models to examine the scope of new doings in market for product • Prepared a business model to enter cold-storage industry with product differentiation & consumer incentivization
Certificate Programme · Data Science and Machine Learning
B.E.
