ML Engineer (Data), Foundational Models at Sarvam AI (Bengaluru)

Job Description

Sarvam is hiring an ML Engineer to own the data infrastructure for its next foundation models in Bengaluru: petabyte-scale curation and filtering pipelines, and the systems that decide what goes into a training run, in what proportion and in what order.

At a glance

• Role: ML Engineer (Data), Foundational Models
• Where: Bengaluru, India
• Experience: 3+ years building large-scale data systems (exceptional early-career candidates considered)
• Focus: pre-training and post-training data pipelines, quality filtering and mixture design
• Apply: open until filled (re-checked 12 October 2026)

About the training data role

• Build pre-training and post-training pipelines at petabyte scale: ingestion, parsing, normalisation, filtering, deduplication, tokenisation and packing
• Develop quality filtering, including model-based quality classifiers and contamination detection
• Own data mixture design, curriculum and annealing strategies with the research team
• Build tools that let researchers analyse, slice, attribute and debug the data
• Scale the pipeline to multilingual corpora, code, maths and multi-source web data

Indian-language AI is a distinct research problem: many languages have little training data, scripts vary widely, and users mix languages in a single sentence. Building models, evaluations and serving systems that work well for these users is what makes roles at an Indian foundation-model company different from similar roles elsewhere.

What Sarvam is looking for

• A BS or MS in computer science or a related field, or equivalent experience
• 3+ years building large-scale data systems such as petabyte-scale or distributed pipelines
• Hands-on data curation and filtering for LLM training, and the ability to defend the choices behind a corpus you built
• Deep familiarity with Spark, Ray, Beam, Dask or similar, and the storage underneath them
• Strong Python, comfort with tokenisation, sharding, packing and IO performance, and open-source contributions in data tooling

Nice to have

• Building or working with large open pre-training corpora
• Multilingual data collection, normalisation, quality scoring and mixing
• Model-based quality classifiers, contamination detection or data attribution
• Tokenisation research, and first-author papers or technical reports

Pay and location

Sarvam’s posting does not state a salary. The role is based in Bengaluru.

How to apply

Apply through the official Sarvam job posting. Sarvam’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 12 October 2026. Check the posting for location and work-authorisation details before you apply.

See all our artificial intelligence jobs, or browse more research jobs on ResearchJobs.in.