Sumitam Sahoo
Data Scientist (AI/ML) at Infosys
- Role
- Data Scientist (Ai Ml) at Infosys
- Location
- Kolkata, WB, IN
- LinkedIn followers
- 500 followers
About Sumitam Sahoo
I am seeking a Good Job role to explore and enhance my Data Science knowledge gained at Infosys Limited in the last 7 years. Working knowledge of NumPy, Pandas, Matplotlib and Machine Learning models | Looking for a role of Data Scientist (AI/ML).
Experience
Data Scientist (Ai Ml)
Feb 2021 — Present
Data Scientist (AI/ML) at Infosys Limited. Working here Since 15th Feb 2021 Project: Loan Default Risk Prediction for Westpac Bank Developed a machine learning pipeline using Random Forest Classifier to predict loan defaults with 15% higher accuracy than previous models. Designed and implemented derived features using Pandas and NumPy, such as historical performance indicators and trend-based metrics. Cleaned and preprocessed data by handling missing values, outliers, and categorical encoding Visualized trends, seasonality, and risk distribution using Matplotlib for business interpretation Generated customer-level risk scores and daily CSV outputs with risk categories and action flags Collaborated with business teams and mentored juniors to ensure domain alignment and model adoption Oracle E-Business Suite Downtime Prediction – Banking Client Used time series models (ARIMA & Auto-ARIMA) to predict when Oracle EBS might face downtime or slow performance, based on system metrics like CPU, I/O, active sessions, and concurrent requests. Measured accuracy with RMSE, MAE, R², precision, and recall, improving early detection of risky periods by 15% compared to manual checks. Built an automated Python + SQL pipeline that converted system metrics into forecasts updated regularly. Helped Apps DBA team schedule patching, manage concurrent programs better, and reduce SLA breaches.Project: Fraud Detection for Citi Bank Managing a machine learning pipeline using Random Forest Classifier to predict fraud defaults with 10% higher accuracy than previous models. Utilized NumPy and Pandas for data preprocessing, and Matplotlib to visualize fraud trends and risk zones. Evaluated models using F1-score, ROC-AUC, and confusion matrix for business-critical recall
Education
Supreme Knowledge Foundation
Bachelor of Technology, Engineering Science
2012 — 2016
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.