Staff, Machine Learning Engineer (L4)

Twilio runs the messaging and voice infrastructure behind a large share of the notifications and OTPs Indians receive every day, and this role builds the machine learning that runs on top of it. The posting frames the job as scoping, designing and deploying machine learning systems into the real world, partnering with product and engineering teams to deliver Twilio's AI and ML roadmap, and owning large scale ML solutions end to end. In practice that means production machine learning rather than research: training and validating both deep learning and statistical models, choosing between them based on the use case, complexity, performance and robustness, then keeping them accurate once real traffic hits. You would work with data platform teams on batch and real time pipelines, build tooling that helps other engineers ship models, and set engineering standards through mentoring and code review. Despite the L4 label this is a senior individual contributor position, and the seven year requirement is real.

What the posting requires

- 7+ years of applied machine learning with strong Python
- A real grounding in machine learning foundations and the building blocks of modern deep learning
- A track record of building, shipping and maintaining models in production in an ambiguous, fast paced environment
- Experience designing large scale experiments and analysis that inform a product roadmap
- Understanding of why frameworks like PyTorch, TensorFlow or Keras work as they do, not just how to call them
- MLOps familiarity: testing, retraining and monitoring models once live
- The ability to ramp up quickly in unfamiliar business domains
- Exposure to modern data storage, messaging and processing tools such as Kafka, Spark, Hadoop, Presto and DynamoDB, including coding against big data components
- Experience working in an agile team with shifting priorities
- Experience working on AWS

Counts as a bonus, not a requirement

- Experience with large language models

What you would actually be doing

- Building and maintaining scalable machine learning in production
- Training and validating deep learning and statistical models, weighing use case, complexity, performance and robustness
- Developing an end to end understanding of the systems and the reasoning behind them
- Partnering with product managers, tech leads and stakeholders to scope what needs building
- Working with data platform teams on robust batch and real time pipelines
- Mentoring, knowledge sharing, and upholding code review, automated testing and monitoring practice

Location and working pattern

Remote and based in India. Unlike Twilio's other two openings today, this one states no restriction to particular states. Occasional travel may be required for in person team meetings.

Also worth knowing

Twilio lists competitive pay, generous time off, parental and wellness leave, healthcare and a retirement savings programme, varying by location.

A good fit if

You have several years of production machine learning and want unusual scale plus a genuinely remote Indian role.

Think twice if

Your experience is mostly notebooks, coursework or research without deployment. The posting returns repeatedly to shipping and maintaining models, not building them once.