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

Hi, I’m SANTHOSH KUMAR REDDY, an AI/ML Engineer and Data Scientist with 5 years of experience building credit risk models, clinical prediction systems, and automation pipelines used by analysts and clinicians. I deliver Python-based solutions across XGBoost, PyTorch, and SQL, and have deployed explainable models on AWS with real-time dashboards and alerts. I also build GenAI workflows using LangChain to convert ML outputs into executive-ready summaries, cutting report prep time and improving clarity. I thrive in cross-functional environments, collaborating with risk, operations, and compliance teams to validate model thresholds and align outputs with regulatory standards. My toolkit spans modeling, data engineering, and operationalization, with a focus on delivering transparent, scalable AI solutions that empower decision-makers.…Hi, I’m SANTHOSH KUMAR REDDY, an AI/ML Engineer and Data Scientist with 5 years of experience building credit risk models, clinical prediction systems, and automation pipelines used by analysts and clinicians. I deliver Python-based solutions across XGBoost, PyTorch, and SQL, and have deployed explainable models on AWS with real-time dashboards and alerts. I also build GenAI workflows using LangChain to convert ML outputs into executive-ready summaries, cutting report prep time and improving clarity. I thrive in cross-functional environments, collaborating with risk, operations, and compliance teams to validate model thresholds and align outputs with regulatory standards. My toolkit spans modeling, data engineering, and operationalization, with a focus on delivering transparent, scalable AI solutions that empower decision-makers.

SANTHOSH KUMAR REDDY

AI Engineer, Data Scientist, AI Developer, +2





Hi, I’m SANTHOSH KUMAR REDDY, an AI/ML Engineer and Data Scientist with 5 years of experience building credit risk models, clinical prediction systems, and automation pipelines used by analysts and clinicians. I deliver Python-based solutions across XGBoost, PyTorch, and SQL, and have deployed explainable models on AWS with real-time dashboards and alerts. I also build GenAI workflows using LangChain to convert ML outputs into executive-ready summaries, cutting report prep time and improving clarity. I thrive in cross-functional environments, collaborating with risk, operations, and compliance teams to validate model thresholds and align outputs with regulatory standards. My toolkit spans modeling, data engineering, and operationalization, with a focus on delivering transparent, scalable AI solutions that empower decision-makers.…Hi, I’m SANTHOSH KUMAR REDDY, an AI/ML Engineer and Data Scientist with 5 years of experience building credit risk models, clinical prediction systems, and automation pipelines used by analysts and clinicians. I deliver Python-based solutions across XGBoost, PyTorch, and SQL, and have deployed explainable models on AWS with real-time dashboards and alerts. I also build GenAI workflows using LangChain to convert ML outputs into executive-ready summaries, cutting report prep time and improving clarity. I thrive in cross-functional environments, collaborating with risk, operations, and compliance teams to validate model thresholds and align outputs with regulatory standards. My toolkit spans modeling, data engineering, and operationalization, with a focus on delivering transparent, scalable AI solutions that empower decision-makers.

Available to hire

Hi, I’m SANTHOSH KUMAR REDDY, an AI/ML Engineer and Data Scientist with 5 years of experience building credit risk models, clinical prediction systems, and automation pipelines used by analysts and clinicians. I deliver Python-based solutions across XGBoost, PyTorch, and SQL, and have deployed explainable models on AWS with real-time dashboards and alerts. I also build GenAI workflows using LangChain to convert ML outputs into executive-ready summaries, cutting report prep time and improving clarity.

I thrive in cross-functional environments, collaborating with risk, operations, and compliance teams to validate model thresholds and align outputs with regulatory standards. My toolkit spans modeling, data engineering, and operationalization, with a focus on delivering transparent, scalable AI solutions that empower decision-makers.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI/ML Engineer at Principal Financial Group

January 1, 2025 - November 18, 2025

Built credit default risk classification models using XGBoost and Logistic Regression on 70K+ account-level records; improved 90-day default prediction accuracy by 18%; engineered 45+ behavioral and transactional features with SQL and Pandas; boosted model AUC by 11% via feature enrichment. Developed a Streamlit internal dashboard exposing model predictions with SHAP explainability, lift charts, and segment-level confidence bands. Integrated OpenAI GPT-4 via LangChain to auto-summarize weekly portfolio reports and convert anomaly flags into executive-ready narrative; reduced report drafting time by 60%. Containerized the ML pipeline with Docker and deployed on AWS EC2 with batch inference and S3 outputs connected to Tableau dashboards. Collaborated with risk strategy, credit ops, and compliance to align outputs with FCRA/FDIC regulatory standards.

Software Engineer II | (Machine Learning Focus) at Optum Global Solutions

January 1, 2024 - January 1, 2024

Built a custom OCR and NER pipeline combining CNN-based image models with SpaCy and domain-specific rules to extract structured data from scanned provider enrollment forms; improved field-level accuracy by 22% and reduced manual correction effort by 40%. Trained time-series models (ARIMA, Linear Regression) to forecast batch job runtimes and system delays; integrated with ServiceNow to auto-suggest root causes in incident logs, saving 6–8 hours/week in triage. Built explainable ML models (XGBoost, Logistic Regression, and LSTM) on provider, claim, and EHR datasets; used SHAP to interpret outputs and improved denial prediction F1 score by 15% after feature refinement. Designed and deployed scalable ETL pipelines using AWS S3, Glue, Redshift, Talend, and Terraform, with CloudWatch monitoring and CI/CD workflows. Set up Grafana dashboards integrated with CloudWatch logs and Spark pipeline metrics to monitor model drift, data quality issues, and ETL health; enabled real-time alerting and

Data Scientist at Sage Softtech

August 1, 2021 - August 1, 2021

Built readmission risk prediction models using Logistic Regression and XGBoost on de-identified patient data; tuned with SMOTE for class imbalance and lifted precision by 21% on top decile risk segments. Conducted survival analysis using Kaplan-Meier curves and Cox Proportional Hazards Model to identify factors influencing post-discharge mortality; results informed clinical review protocols across 2 regional units. Ran A/B tests with stratified sampling and propensity score matching to measure impact of care-coordination programs on ER revisit rates; surfaced a statistically significant 9% reduction (p < 0.05) in the treatment group. Built and interpreted unsupervised models (K-Means, DBSCAN) to segment patient populations for targeted intervention, identifying 3 high-risk cohorts.