Harssh Verma — Data Scientist

​I build end-to-end machine learning systems that solve expensive business problems. I bridge the gap between traditional Data Analytics and modern Applied AI. I have 3 years of experience deploying models, automating workflows, and optimizing data pipelines. ​My impact by the numbers: ​Engineering Efficiency: Automated portfolio auditing at Optimum. I reduced manual cycle time from 36 hours to under 2 minutes using Python and SQL. ​Data Quality: Led Workday HCM conversions at Deloitte. I improved first-build error rates from 25% to 2% by designing robust SQL validation scripts. ​Technical Scale: Migrated over 1 million records with 98% accuracy across banking and healthcare sectors. ​My technical focus: ​Applied AI & LLMs: I build Retrieval-Augmented Generation (RAG) systems using LangChain, Pinecone, and LLaMA 3. I automate processes using local LLMs (Ollama) and n8n. ​Machine Learning Ops: I deploy calibrated classifiers using Docker and FastAPI. I focus on explainability (SHAP) and model evaluation. ​Analytics: I serve actionable insights using Power BI, DAX, and advanced SQL. ​Core Toolkit: Python, SQL, Docker, AWS (Lambda, S3), Git, RAG, Scikit-learn, XGBoost, Streamlit.

Stackforce AI infers this person is a Data Scientist specializing in Applied AI and Business Intelligence.

Location: Delhi, India

Experience: 2 yrs 3 mos