Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Sai Karthik. I’m a Data Scientist with around four years of experience delivering AI, ML, and data-driven solutions for finance and retail. I love turning complex data into clear, actionable insights using Python, SQL, PyTorch, TensorFlow, XGBoost, LSTM, and MLflow. Over the years I’ve built credit-risk models for 180,000+ loan accounts, deployed ML APIs handling 150,000+ requests per day, and improved forecasting and anomaly detection across millions of transactions. I enjoy collaborating with cross-functional teams and applying LLM pipelines, GPT-4, LangChain, and vector databases to optimize business operations and outcomes.…Hi, I’m Sai Karthik. I’m a Data Scientist with around four years of experience delivering AI, ML, and data-driven solutions for finance and retail. I love turning complex data into clear, actionable insights using Python, SQL, PyTorch, TensorFlow, XGBoost, LSTM, and MLflow. Over the years I’ve built credit-risk models for 180,000+ loan accounts, deployed ML APIs handling 150,000+ requests per day, and improved forecasting and anomaly detection across millions of transactions. I enjoy collaborating with cross-functional teams and applying LLM pipelines, GPT-4, LangChain, and vector databases to optimize business operations and outcomes.

Sai Karthik

Data Scientist, AI Engineer, AI Developer, +2





Hi, I’m Sai Karthik. I’m a Data Scientist with around four years of experience delivering AI, ML, and data-driven solutions for finance and retail. I love turning complex data into clear, actionable insights using Python, SQL, PyTorch, TensorFlow, XGBoost, LSTM, and MLflow. Over the years I’ve built credit-risk models for 180,000+ loan accounts, deployed ML APIs handling 150,000+ requests per day, and improved forecasting and anomaly detection across millions of transactions. I enjoy collaborating with cross-functional teams and applying LLM pipelines, GPT-4, LangChain, and vector databases to optimize business operations and outcomes.…Hi, I’m Sai Karthik. I’m a Data Scientist with around four years of experience delivering AI, ML, and data-driven solutions for finance and retail. I love turning complex data into clear, actionable insights using Python, SQL, PyTorch, TensorFlow, XGBoost, LSTM, and MLflow. Over the years I’ve built credit-risk models for 180,000+ loan accounts, deployed ML APIs handling 150,000+ requests per day, and improved forecasting and anomaly detection across millions of transactions. I enjoy collaborating with cross-functional teams and applying LLM pipelines, GPT-4, LangChain, and vector databases to optimize business operations and outcomes.

Available to hire

Hi, I’m Sai Karthik. I’m a Data Scientist with around four years of experience delivering AI, ML, and data-driven solutions for finance and retail. I love turning complex data into clear, actionable insights using Python, SQL, PyTorch, TensorFlow, XGBoost, LSTM, and MLflow.

Over the years I’ve built credit-risk models for 180,000+ loan accounts, deployed ML APIs handling 150,000+ requests per day, and improved forecasting and anomaly detection across millions of transactions. I enjoy collaborating with cross-functional teams and applying LLM pipelines, GPT-4, LangChain, and vector databases to optimize business operations and outcomes.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI Data Scientist at BlackRock

June 1, 2024 - Present

Engineered AI-powered portfolio risk monitoring pipelines processing 80K+ client portfolios monthly; generated automated risk alerts reducing manual review time by 6 hours per analyst weekly. Built NLP-driven document workflows with GPT-4, Hugging Face Transformers, LangChain, and vector databases to summarize and classify 60K+ loan/investment/regulatory documents monthly; reduced manual review time from 7 hours to under 2 hours. Developed anomaly detection models using Isolation Forest and autoencoders, improving detection coverage across 120K+ monthly client interactions. Created client segmentation and clustering models to identify high-value and high-risk segments for 35K+ portfolios. Operationalized end-to-end ML pipelines via Airflow and MLflow for 20+ production models, reducing data latency. Conducted predictive modeling for retention and investment recommendations; engaged 2,100+ at-risk clients per quarter; improved rebalancing speed by 1.8x. Executed end-to-end A/B testing,

Data Scientist at KPIT Technologies

February 1, 2020 - July 1, 2022

Architected credit-risk scoring models using Python, XGBoost, logistic regression, and SHAP to evaluate repayment behavior for 180,000+ loan accounts, enabling risk officers to flag 7,200 high-risk profiles for early intervention. Formulated demand forecasting and inventory optimization models using LSTM, RNN, and GRU networks to predict sales for 5,000+ SKUs across 200+ retail outlets, reducing stockouts by 1,200 units and tripling replenishment speed. Devised NLP pipelines to classify client emails, KYC documents, and regulatory disclosures, reducing manual review time by 5 hours per 1,000 documents. Constructed anomaly detection for 7 million monthly transactions, helping intercept 2,300+ suspicious activities. Designed customer segmentation and portfolio clustering for 420,000+ client accounts; enabled targeted campaigns with 8,500+ high-value engagements. Deployed real-time ML APIs using Docker, FastAPI, and MLflow for loan approvals and fraud detection, handling 150,000+ requests