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

AI/ML Engineer with nearly 6 years of experience architecting and deploying enterprise-scale GenAI, NLP, and predictive modeling solutions. Built RAG pipelines, fine-tuned domain-specific models, and developed MLOps workflows that reduced ticket resolution time by 40% and cut cloud serving costs by $120K. Experienced in high-throughput streaming data and end-to-end model lifecycles using Apache Spark, Kafka, Azure ML, and Kubernetes, while driving business impact like reducing customer churn by 18%. Strong advocate for responsible AI with bias detection, explainability (SHAP/LIME), and monitoring for enterprise governance.…AI/ML Engineer with nearly 6 years of experience architecting and deploying enterprise-scale GenAI, NLP, and predictive modeling solutions. Built RAG pipelines, fine-tuned domain-specific models, and developed MLOps workflows that reduced ticket resolution time by 40% and cut cloud serving costs by $120K. Experienced in high-throughput streaming data and end-to-end model lifecycles using Apache Spark, Kafka, Azure ML, and Kubernetes, while driving business impact like reducing customer churn by 18%. Strong advocate for responsible AI with bias detection, explainability (SHAP/LIME), and monitoring for enterprise governance.

Praveen Kumar

AI Engineer, Data Scientist, AI Developer, +5





AI/ML Engineer with nearly 6 years of experience architecting and deploying enterprise-scale GenAI, NLP, and predictive modeling solutions. Built RAG pipelines, fine-tuned domain-specific models, and developed MLOps workflows that reduced ticket resolution time by 40% and cut cloud serving costs by $120K. Experienced in high-throughput streaming data and end-to-end model lifecycles using Apache Spark, Kafka, Azure ML, and Kubernetes, while driving business impact like reducing customer churn by 18%. Strong advocate for responsible AI with bias detection, explainability (SHAP/LIME), and monitoring for enterprise governance.…AI/ML Engineer with nearly 6 years of experience architecting and deploying enterprise-scale GenAI, NLP, and predictive modeling solutions. Built RAG pipelines, fine-tuned domain-specific models, and developed MLOps workflows that reduced ticket resolution time by 40% and cut cloud serving costs by $120K. Experienced in high-throughput streaming data and end-to-end model lifecycles using Apache Spark, Kafka, Azure ML, and Kubernetes, while driving business impact like reducing customer churn by 18%. Strong advocate for responsible AI with bias detection, explainability (SHAP/LIME), and monitoring for enterprise governance.

Available to hire

AI/ML Engineer with nearly 6 years of experience architecting and deploying enterprise-scale GenAI, NLP, and predictive modeling solutions. Built RAG pipelines, fine-tuned domain-specific models, and developed MLOps workflows that reduced ticket resolution time by 40% and cut cloud serving costs by $120K.

Experienced in high-throughput streaming data and end-to-end model lifecycles using Apache Spark, Kafka, Azure ML, and Kubernetes, while driving business impact like reducing customer churn by 18%. Strong advocate for responsible AI with bias detection, explainability (SHAP/LIME), and monitoring for enterprise governance.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI/ML Engineer at ServiceNow

November 1, 2024 - Present

Architected and deployed LLM-powered automation workflows on the ServiceNow Now Platform, reducing manual ticket resolution time by 40% across IT and HR service desks. Built and fine-tuned domain-specific NLP models for intelligent incident classification and routing, achieving 92% accuracy on multi-class categorisation at enterprise scale. Integrated Generative AI (RAG pipelines) with ServiceNow's Virtual Agent, enabling context-aware self-service resolution and deflecting 30% of Tier-1 support volume. Designed scalable MLOps pipelines using Azure ML and ServiceNow Flow Designer, automating model retraining, versioning, and A/B testing in production environments. Collaborated cross-functionally with platform engineers and product managers to embed predictive analytics dashboards inside ServiceNow ITSM, surfacing real-time SLA breach risk scores. Championed responsible AI practices by implementing bias detection, explainability (SHAP/LIME), and monitoring frameworks, ensuring complianc

Machine Learning Engineer at Neon IT Systems

January 1, 2020 - July 1, 2023

Led end-to-end development of a customer churn prediction system using XGBoost and deep neural networks, reducing annual churn by 18% for a telecom client with 2M+ subscribers. Engineered real-time recommendation engines using collaborative filtering and matrix factorisation, increasing an e-commerce client's average order value by 22%. Designed and maintained robust data pipelines using Apache Spark and Kafka, processing 50GB+ of streaming data daily with sub-second latency for fraud detection use cases. Implemented computer vision solutions (ResNet, YOLO) for automated quality inspection in a manufacturing client, reducing defect escape rate by 35% and inspection labour costs by 25%. Established team-wide MLOps best practices — standardized experiment tracking with MLflow, containerised model serving via Docker/Kubernetes, and CI/CD model deployment pipelines. Mentored a team of 4 junior ML engineers, conducting weekly code reviews and technical sessions on model optimisation, feat

Data Scientist at Green Touch Technology

November 1, 2018 - December 1, 2019

Conducted exploratory data analysis and statistical modelling on large structured datasets, delivering actionable insights that informed product pricing strategy and increased margin by 12%. Built sales forecasting models using ARIMA and Prophet, achieving 91% accuracy over 30-day horizons and enabling data-driven inventory planning for retail clients. Developed interactive business intelligence dashboards using Tableau and Python (Plotly/Dash), enabling non-technical stakeholders to self-serve insights and reducing ad-hoc reporting requests by 50%. Performed feature engineering and model selection across classification and regression tasks (Random Forest, SVM, logistic regression), consistently improving baseline model performance by 15–20%. Automated ETL workflows using Python and SQL, cutting manual data preparation time from 8 hours to under 30 minutes per reporting cycle and improving data quality for downstream analytics.