Shashank Garg

Data Scientist @Pragmatic Semiconductor

Cambridge, GB

Driving cross-functional data science initiatives that raise product quality, automate operations, and scale ML/LLM solutions across engineering and business workflows.• Optimised operations: built and deployed ML & DL models (Python · TensorFlow · XGBoost) that improved Key Performance Indicators (KPIs) such as efficiency, defect rates and ageing metrics.• LLM at scale: designed a Retrieval-Augmented Generation (RAG) chatbot with OpenAI API, Anthropic API & LangChain; cut support-query response times and freed team capacity. Deployed via FastAPI on AWS EC2.• Anomaly detection: engineered time-series signal-detection pipelines that flag performance drift early, reducing manual inspection errors.• Real-time insights: created interactive Tableau dashboards fed by Python scripts and SQL pipelines that surface error rates, product performance and team efficiency for R&D and Ops.• Data pipelines: authored SQL-based workflows on AWS (EC2, S3) to automate preprocessing and boost data accessibility and reporting accuracy.• Stakeholder alignment: partnered with engineering, operations and business leads to ensure every data solution is actionable, scalable and tied to strategic goals.Tech Stack:• Programming: Python, R, SQL• LLM & NLP: OpenAI API, Anthropic API, LangChain, Hugging Face Transformers, Google Cloud NLP API• Vector DBs: FAISS, Pinecone, Chroma• ML / DL: scikit-learn, TensorFlow, XGBoost, PyTorch, Keras• Data Processing: pandas, NumPy, SciPy, statsmodels• Visualisation: Tableau, Power BI, Matplotlib, Seaborn• Cloud / DevOps: AWS (EC2, S3, SageMaker), Docker, FastAPI• Environments: VS Code, PyCharm, Jupyter Notebook

University of Exeter

Masters

G. B. Pant University Of Agriculture & Technology

Bachelor of Technology