Sudhanva Narayana

Professional summary

Sudhanva Narayana is a Senior Machine Learning Engineer in the San Francisco Bay Area specializing in production ML systems, infrastructure, inference, model serving, and cloud-native AI platforms. Current work includes billion-row TensorFlow prediction pipelines, Kubernetes platforms with Ray and Flyte, multi-GPU model delivery, observability, and cloud cost optimization.

Experience

1. Jul 2023 – Present

   Remote from the San Francisco Bay Area · employer based in Boston, MA

   Senior Machine Learning Engineer · Montai Therapeutics

   • Built batch prediction pipelines processing 1B+ rows across TensorFlow models in under three hours, accelerating drug-candidate evaluation and experimental iteration.
   • Owned end-to-end ML infrastructure: autoscaling Kubernetes with Ray and Flyte, observability with Prometheus and Grafana, alerting, and more than $50,000 in annual cloud-cost reduction.
   • Architected a multi-GPU build system that reduced model deployment time by 50%; automated tuning and CI/CD saved about ten engineering hours each week.
   • Developed internal applications that let cross-functional teams run models on demand with experiment tracking and model versioning.

   Billion-Row Batch Inference case study → Kubernetes ML Platform case study → Multi-GPU Deployment System case study →

2. May 2022 – Jul 2022

   San Francisco, CA

   Machine Learning Engineer Intern · Autodesk

   • Deployed a high-throughput transformer on a real-time event pipeline, analyzing more than one million daily user interactions for next-click prediction.

   Real-Time Transformer Serving case study →

3. Jun 2020 – May 2021

   Bengaluru, India

   Machine Learning Engineer · Pixxel

   • Built multi-regional ML infrastructure for geospatial analytics, improving inference efficiency by 75% and reducing map-rendering latency by 50%.
   • Built ETL and ML pipelines processing more than 1 TB each day and reduced pipeline failures from data-schema issues by 30%.
   • Architected a GPU queue with automated A/B testing for regional model selection, saving $50,000 annually in cloud costs.

   Geospatial ML Infrastructure case study →

4. May 2019 – May 2020

   Bengaluru, India

   Machine Learning Engineer · Initiable Intelligence

   • Led a team of four to build an NLP chatbot API that enabled five-minute integration for static sites and was adopted by multiple startups.

5. Jan 2017 – May 2019

   Bengaluru, India

   Software Engineer · DCT Academy

   • Built a task recommendation engine with automated assignment matching, improving student engagement by 60%.
   • Built a classroom-management tool with auto-grading and student clustering, saving about two hours each day for teaching staff.

Education

• Sep 2021 – May 2023

  Master of Science in Artificial Intelligence

  Northeastern University · Boston, MA

• Jun 2013 – Nov 2016

  Bachelor of Science in Computer Science

  PES University · Bengaluru, India

Technical capabilities

ML infrastructure

Kubernetes · Ray Data and Serve · Flyte · Airflow · Terraform · Argo CD

Models and data

TensorFlow · PyTorch · scikit-learn · pandas · Milvus · MLflow

Reliability

Prometheus · Grafana · alerting · GitHub Actions · Docker · GitOps

Languages and cloud

Python · SQL · AWS · GCP · Azure · Oracle Cloud