AI/ML Engineer with Real-World Impact

TARAKA RAMA NARASIMHA AI/ML Engineer

****************************@*****.*** +1-904-***-**** LinkedIn

PROFESSIONAL SUMMARY

AI/ML Engineer with 3+ years of experience building scalable and genuinely useful machine learning solutions across diverse industries. I love turning messy data into clean, reliable predictive systems using tools like Python, TensorFlow, PyTorch, and Scikit-learn. I’m known for blending solid engineering discipline with a creative problem-solver vibe, consistently boosting model performance and streamlining pipelines. I thrive in cross-team collabs, working closely with product and engineering partners to deliver ML features that actually move metrics. Always excited about crafting systems that make everyday workflows smoother and help organizations level up through data-driven decisions.

TECHNICAL SKILLS

Programming: Python, SQL, Java, Scala

Machine Learning: Scikit-learn, TensorFlow, PyTorch, Keras, XGBoost, LightGBM

Deep Learning: CNNs, RNNs, Transformers, BERT, GPT APIs

NLP: spaCy, Hugging Face, NLTK

Data Engineering: Airflow, Databricks, Spark, Kafka, Feast

Cloud Platforms: AWS (SageMaker, Lambda, S3, EC2), Azure ML

MLOps: Docker, Kubernetes, MLflow, CI/CD (GitHub Actions, Jenkins)

Databases: PostgreSQL, MySQL, MongoDB, BigQuery

Tools: Git, Grafana, JIRA, Elastic Stack, FastAPI

Project Management: Agile/Scrum, User Story Mapping, Cross-functional Collaboration

PROFESSIONAL EXPERIENCE

FREYR SOLUTIONS – Princeton, NJ

Machine Learning Engineer Oct 2024 – Present

Designed and deployed ML-based ticket-routing models using Python, Scikit-learn, and XGBoost, improving support resolution speed by 35% across enterprise service desks.

Developed advanced intent-classification pipelines leveraging spaCy, Hugging Face Transformers, and BERT, boosting chatbot response accuracy by 25%.

Migrated existing training workloads to Azure Machine Learning, enabling scalable distributed training and reducing compute overhead by 20%.

Built automated CI/CD workflows for ML packaging and deployment using Jenkins, Git, and MLflow, cutting release cycles by 50%.

Developed real-time analytics dashboards with Elastic Stack, raising incident-detection speed by 40% and improving SLA compliance.

Optimized model inference with ONNX Runtime, achieving a 3x increase in throughput for high-volume prediction endpoints.

Automated data-quality and validation pipelines using Python, Pandas, and Spark, removing manual QA steps and saving ~10 hours weekly.

Partnered with architects and PMs to shape the ML strategy, ensuring alignment with enterprise KPIs and long-term platform goals.

Implemented proactive monitoring and alerting for ML pipeline failures using Grafana, improving system reliability and reducing downtime by 60%.

Delivered explainable AI reports using SHAP and LIME, strengthening stakeholder trust and supporting compliance-driven decision making.

BANK OF AMERICA – Charlotte, NC

AI/ML Engineer Oct 2023 – Sep 2024

Built fraud-detection models using XGBoost and LightGBM, improving detection accuracy by 28% while reducing false positives across high-risk transactions.

Supported development of credit-risk scoring models for 500K+ customer applications, helping strengthen approval and underwriting workflows.

Improved data-quality and ETL pipelines using Spark and Databricks, cutting end-to-end processing time by 45%.

Set up model-drift monitoring dashboards and alerts, reducing model-decay incidents by 20% in production.

Automated feature-generation pipelines using Feast, enabling faster experimentation across multiple model iterations.

Ran A/B experiments for new ML model versions, contributing to a 12% boost in fraud-catch performance.

Enhanced real-time inference APIs developed with FastAPI, lowering latency to under 100 ms for risk checks.

Deployed ML models on AWS SageMaker, improving operational stability and meeting banking compliance requirements.

Created model documentation and compliance reports aligned with audit standards, reducing review time by 30%.

Collaborated with data engineering teams to maintain consistent data schemas and improve reliability across risk pipelines.

VIVAGENE – Hyderabad, India

Machine Learning Engineer Aug 2021 – Nov 2022

Helped build deep-learning models for medical image classification using TensorFlow, PyTorch, and CNNs, improving prediction accuracy by 22%.

Supported development of patient-risk scoring models that reduced manual triage work by 30% and improved care team prioritization.

Created data processing workflows in Azure ML, Azure Storage, and Azure Functions, improving system reliability and reducing manual handling.

Assisted in developing anomaly-detection pipelines for EHR activity using Scikit-learn, reaching 92% precision in early prototypes.

Built NLP components using BERT and Hugging Face, speeding up clinical note review processes by 40%.

Improved ETL tasks with Airflow, reducing data refresh cycles from hours to minutes and supporting faster model retraining.

Worked with clinical analysts to test model outputs and contributed to an 18% reduction in false positives through iterative feedback cycles.

Containerized ML workloads with Docker and supported deployment on Kubernetes, improving rollout consistency by 35%.

Set up monitoring dashboards in Grafana to track pipeline performance and reduce debugging time for engineering teams.

Contributed to CI/CD automation in GitHub Actions, enabling reproducible training runs and lowering release-related issues by 25%.

EDUCATION

Master’s in Information Technology – Webster University

Contact this candidate