Staff MLOps Engineer · Engineering Lead, MLOps · Author
I make machine learning land in production, and keep it running.
👋 I'm Prafful Mishra, a Staff MLOps & Platform Engineer based in Sweden. I operationalise ML by making sure the cool stuff lands in production and gets maintained, bridging the gap between ML practitioners, software engineers, and SREs.
🔭 Talk to me about
- Federated Learning
- Ethical AI
- MLOps
- Planes
- Cars
- Kung Fu Panda
- Big Hero 6
- Bruce Wayne
- Batman
- Gotham
About
I'm a Staff MLOps Engineer and engineering manager who is happiest at the seams of a system, where data, models, infrastructure, and the people building them all have to agree. I lead the MLOps engineering team at Epidemic Sound, building robust, maintainable ML platforms that let teams deploy and monitor models with confidence, and coaching the engineers who keep them running.
In 2025 I published “A Guide to Implementing MLOps: From Data to Operations” with Springer, a practical walk through building end-to-end MLOps pipelines. I also write regularly on Medium and Substack about MLOps, platforms, and the messy realities of running ML in production.
I hold a Bachelors of Engineering (Honours) in Computer Science & Engineering from Rajiv Gandhi Prodyogiki Vishwavidyalaya, India. My research interests sit in Federated Learning and Quantum Computing.
Research publications
- Accuracy Crawler: An Accurate Crawler for deep web data harvesting
- A Probabilistic Weighted Ensemble Algorithm
Featured publication
A Guide to Implementing MLOps: From Data to Operations
A comprehensive, practical guide to building end-to-end MLOps pipelines, from data versioning through deployment to monitoring. Written for engineers who want production-grade ML without the hand-waving, drawn from real-world platform work.
Read on Springer →Work
-
Staff MLOps Engineer · Engineering Lead
Leading the MLOps engineering team, and the platform it builds.
- Managing the team: weekly 1:1s, growth and learning plans, career coaching, and quarterly priorities.
- Running the goalie/on-call rotation, the hiring loop, and peer-team interview panels.
- Representing the team's roadmap to product, ML, and platform leadership.
- Leading the infrastructure for training foundational models and serving the product's GenAI inference.
- Architecting deployment strategies for ML products and optimising cost across ML initiatives.
-
Senior MLOps Engineer
Drove the MLOps roadmap end to end.
- Defined the MLOps roadmap around the organisation's needs.
- Consulted ML engineers on development & production deployment strategy.
- Built and maintained training/serving infrastructure and data pipelines.
-
Machine Learning Engineer
Core member of the central ML Engineering & Operations team.
- Built and ran a Kubernetes-native, multi-cluster central data science platform used across the org.
- Consulted teams across the full lifecycle, from data collection to scalable deployments.
- Unblocked teams to minimise time-to-production for models.
-
Machine Learning Engineer
- Developed, deployed, and maintained ML models across distributed architectures with Kubernetes, Kubeflow, PyTorch, and AWS.
-
Senior Software Developer, Machine Learning
- Researched and helped design architecture for ML solutions within the Applied R&D team.
Stack
The tools and ideas I reach for most.
MLOps & Infrastructure
Machine Learning
Programming & Data
Exploring
Open source
I contribute back to the MLOps and infrastructure tools I build on.
lakeFS
Git-like version control for object storage and data lakes.
View on GitHub → k8sKubeflow
The machine learning toolkit for Kubernetes.
View on GitHub → dagKubeflow Pipelines
Building and deploying portable, scalable ML workflows.
View on GitHub → pyKubernetes Python Client
The official Python client library for the Kubernetes API.
View on GitHub →Writing
Notes from the field, on Medium & Substack.
How to Version Unstructured Data?
Strategies and tooling for versioning unstructured data inside ML pipelines.
Read → May 2023Cracking the Code of Data Science Team Structures
How to build effective data science teams and the org structures around them.
Read → Apr 2023Why Being a Software Engineer is a Competitive Sport
Reflections on the competitive nature of engineering and continuous learning.
Read → Jul 2022Deleting my First Production Cluster
Lessons from managing production infra and why careful operations matter.
Read → Feb 2021Conditional Kubeflow Pipeline for Dummies
A beginner-friendly guide to building conditional pipelines in Kubeflow.
Read → Nov 2019A Dummies' Guide to Build a Kubeflow Pipeline
A step-by-step tutorial on creating your first Kubeflow pipeline from scratch.
Read →Get in touch
Always happy to talk MLOps, platforms, responsible AI, or Kung Fu Panda 🐼. Have a project, a question, or just want to say hi?
hello@mishraprafful.com