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

I am a machine learning engineer with 4+ years of experience delivering production-grade models across classification, forecasting, NLP, and anomaly detection. I’m proficient in Python, SQL, Scikit-learn, TensorFlow, and PyTorch, with hands-on expertise in AUC, precision-recall, and SHAP-based evaluations. I’ve built automated training pipelines, implemented A/B testing, and reduced inference time using containerized deployments.\n\nI design scalable solutions for real-time and batch data pipelines, expose models via REST APIs using Docker, and ensure reliability through drift detection and scheduled retraining to maintain stable performance across diverse production environments.…I am a machine learning engineer with 4+ years of experience delivering production-grade models across classification, forecasting, NLP, and anomaly detection. I’m proficient in Python, SQL, Scikit-learn, TensorFlow, and PyTorch, with hands-on expertise in AUC, precision-recall, and SHAP-based evaluations. I’ve built automated training pipelines, implemented A/B testing, and reduced inference time using containerized deployments.\n\nI design scalable solutions for real-time and batch data pipelines, expose models via REST APIs using Docker, and ensure reliability through drift detection and scheduled retraining to maintain stable performance across diverse production environments.

Chiru Anand Vaka

AI Engineer, Data Scientist, AI Developer, +2





I am a machine learning engineer with 4+ years of experience delivering production-grade models across classification, forecasting, NLP, and anomaly detection. I’m proficient in Python, SQL, Scikit-learn, TensorFlow, and PyTorch, with hands-on expertise in AUC, precision-recall, and SHAP-based evaluations. I’ve built automated training pipelines, implemented A/B testing, and reduced inference time using containerized deployments.\n\nI design scalable solutions for real-time and batch data pipelines, expose models via REST APIs using Docker, and ensure reliability through drift detection and scheduled retraining to maintain stable performance across diverse production environments.…I am a machine learning engineer with 4+ years of experience delivering production-grade models across classification, forecasting, NLP, and anomaly detection. I’m proficient in Python, SQL, Scikit-learn, TensorFlow, and PyTorch, with hands-on expertise in AUC, precision-recall, and SHAP-based evaluations. I’ve built automated training pipelines, implemented A/B testing, and reduced inference time using containerized deployments.\n\nI design scalable solutions for real-time and batch data pipelines, expose models via REST APIs using Docker, and ensure reliability through drift detection and scheduled retraining to maintain stable performance across diverse production environments.

Available to hire

I am a machine learning engineer with 4+ years of experience delivering production-grade models across classification, forecasting, NLP, and anomaly detection. I’m proficient in Python, SQL, Scikit-learn, TensorFlow, and PyTorch, with hands-on expertise in AUC, precision-recall, and SHAP-based evaluations. I’ve built automated training pipelines, implemented A/B testing, and reduced inference time using containerized deployments.\n\nI design scalable solutions for real-time and batch data pipelines, expose models via REST APIs using Docker, and ensure reliability through drift detection and scheduled retraining to maintain stable performance across diverse production environments.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

Work Experience

ML Engineer at UBER

December 1, 2024 - Present

Designed a churn classification model trained on over 5 million ride histories to flag users likely to exit 3 weeks in advance, improving retention by 12% in North America. Built a rules-free surge forecasting system using stacked ensembles and live trip data, resulting in 9% volume growth during peak demand without manual adjustments. Developed anomaly detection pipelines learning weekly pattern shifts, reducing fraud flagging by 27% and saving 230+ operations hours quarterly. Accelerated deployment time from 4 weeks to 5 days by containerizing workflows linked with event-based API endpoints across 11 backend microservices. Implemented A/B testing across 4 user clusters for validating model output differences with confidence bands. Reduced model drift by 35% via region-specific retraining and automated feature drift monitoring, enhancing prediction consistency across 100+ city-level deployments worldwide.

ML Engineer at Citius Tech

July 1, 2023 - August 8, 2025

Trained time-series models on 50,000+ longitudinal patient records to track disease risk stages, reducing chronic case misclassification by 18% across hospitals. Predicted 30-day readmissions with 84% validation accuracy, cutting projected readmission costs by $500K quarterly and improving care cycles. Designed NLP pipelines processing radiology narratives, improving diagnostic overlap by 22% and supporting 5 clinical departments. Modeled chronic care decisions using regression and guideline-derived rules, enhancing enrollment relevance for 30,000+ active cardiovascular and endocrine patients. Reduced inference time by 46% through ONNX model conversion and dockerized inference servers, decreasing radiologist wait times. Developed dashboards to monitor 11 ML model variants, enabling autonomous data team audits without manual rechecks.

ML Engineer at Mphasis

April 30, 2022 - August 8, 2025

Built a binary classifier on 1.2 million loan records increasing default prediction accuracy by 16%, aiding finance teams in reconfiguring approval margins for high-risk segments. Automated feature selection using correlation filters and SHAP, reducing model development time by nearly 40% during critical sprints. Developed sentiment models on 50,000+ customer transcripts classifying resolution tone, leading to 21% fewer ticket escalations and shorter first-call resolution times. Forecasted regional telecom staffing needs using Prophet and ARIMA models, saving $280K annually by optimizing workforce schedules. Migrated scripts to AWS EC2 for reproducible training with orchestrated jobs, enabling consistent performance benchmarks. Created biweekly model drift visualizations and trend summaries for business alignment on inference stability.

ML Engineer at UBER

May 31, 2025 - August 22, 2025

Led the design and deployment of a churn classification model that used over 5 million ride histories to predict user exit three weeks in advance, enhancing retention by 12% across North America. Developed a rule-free surge forecasting system leveraging stacked ensembles and live trip data, realizing a 9% growth in metro area ride volume. Built anomaly detection pipelines to monitor weekly pattern shifts, reducing fraud flagging time by 27% and saving over 230 operations hours quarterly. Accelerated deployment lag drastically by containerizing workflows and integrating event-based APIs across 11 microservices. Implemented A/B testing across multiple user clusters to validate model outcomes with precision. Decreased model drift by 35% through region-specific retraining and automatic feature drift detection, stabilizing predictions for over 100 city deployments.

ML Engineer at Citius Tech

July 31, 2023 - August 22, 2025

Trained time-series models on 50,000+ longitudinal patient records, improving disease risk tracking and reducing misclassification by 18% across multiple hospitals. Predicted 30-day readmissions on EHR data with high accuracy, cutting projected readmission costs by $500K quarterly. Developed NLP pipelines for processing radiology reports, enhancing diagnostic overlap by 22% across five departments. Modeled chronic care decision logic using regression and rules, improving enrollment for over 30,000 active cases. Reduced inference latency by 46% using ONNX model conversions and Docker-based inference servers. Created dashboards to track ML model variants and enable autonomous accuracy audits.

ML Engineer at Mphasis

April 30, 2022 - August 22, 2025

Built a binary loan default classifier on 1.2 million records, boosting prediction accuracy by 16% and aiding finance teams in risk assessment. Automated feature selection with correlation and SHAP, reducing model development time by nearly 40%. Trained sentiment models on 50,000+ customer transcripts to classify resolution tones, leading to a 21% reduction in ticket escalations. Forecasted staffing needs with Prophet and ARIMA, generating $280K annual savings. Migrated legacy scripts to AWS EC2 for reproducible training workflows, supporting multiple in-house ML deployments. Delivered biweekly model drift visualizations facilitating strategic roadmap alignment.

ML Engineer at UBER

December 1, 2024 - November 7, 2025

Designed a churn classification model educated on 5M+ ride histories that flagged exit-prone users 3 weeks in advance, improving retention by 12% after rollout in North America. Built a rules-free surge forecasting system using stacked ensembles and live trip data, leading to 9% volume growth in metro areas during peak demand with zero manual adjustments. Deployed anomaly detection pipelines instructed on weekly pattern shifts, reducing fraud flagging time by 27% and saving 230+ operations hours per quarter for compliance audit teams. Reduced deployment lag from 4 weeks to 5 days by containerizing model workflows and linking them with event-based API endpoints used across 11 backend microservices. Rolled out an A/B testing setup across 4 active user clusters, enabling business review teams to validate daily model output differences using consistent confidence bands and test thresholds. Lowered model drift by 35% through region-specific retraining schedules and automated feature drift c

ML Engineer at Citius Tech

July 1, 2023 - July 1, 2023

Trained time-series models on 50K+ longitudinal patient records to track disease risk stages, helping clinicians reduce misclassification in chronic cases by 18% across hospitals. Predicted 30-day readmissions on structured EHR data with 84% validation accuracy, reducing projected quarterly readmission costs by $500K and improving care cycle alignment. Designed an NLP pipeline that processed radiology narratives, enhancing specialty-specific diagnostic overlap by 22% and supporting 5 departments with error-traced prediction outputs. Modeled chronic care decisions using regression logic and guideline-derived rules, improving enrollment relevance for 30K+ active cases in cardiovascular and endocrine care programs. Compressed inference time by 46% via ONNX-based model conversion and Dockerized inference servers, lowering radiologist wait times in production dashboards. Developed internal dashboards to track 11 ML model variants, allowing data teams to self-audit accuracy shifts and perfor

ML Engineer at Mphasis

April 1, 2022 - April 1, 2022

Built a binary classifier on 1.2M+ loan records that raised default prediction accuracy by 16%, helping the finance team reconfigure approval margins on high-risk segments. Automated feature selection process using correlation filters and SHAP scores, which brought down model development time by nearly 40% across 3 critical delivery sprints. Trained a sentiment model on 50K+ customer transcripts to classify resolution tone, resulting in a 21% drop in ticket escalations and shorter first-call resolution times. Forecasted telecom staffing needs by region using Prophet and ARIMA models, producing $280K/year in budget savings by adjusting workforce schedules in real time. Transitioned legacy scripts to AWS EC2-based training modules and set up reproducible job orchestration, enabling repeatable performance benchmarks for 5 in-house ML deployments. Prepared biweekly model drift visualizations and trend summaries for business heads, aligning quarterly roadmap decisions with underlying infere