Senior ML Engineer

Eka.Care; Bengaluru; On-site; INR 3,000,000 - 5,000,000; Full time

Benefits offered by this job

Medical Insurance & Accidental

Job summary

EkaCare is seeking a versatile software engineer with ML training exposure to advance our large-scale model training and inference pipelines in Bengaluru. You will own data pipelines, distributed training workflows, and evaluation tooling for models reaching billions of parameters.

You will work on Spark/Ray data workflows, Kubernetes and Slurm orchestration, and build robust, observable systems that ship publicly with open-source contributions when possible.

Qualifications

• 3–6 years of strong software engineering with real ML training exposure.
• You’ve fine-tuned or trained ≥1B-parameter models, not just called APIs.
• Distributed computing fundamentals: Slurm or Kubernetes, NCCL basics, storage/throughput intuition.
• Data engineering at scale; Spark, Ray, or whatever gets the job done.
• A bias for ownership: you don’t ask whose job it is.

Responsibilities

• Build corpus pipelines for large-scale data prep and ingestion.
• Run distributed training jobs on multi-node clusters; monitor, recover, and optimize.
• Develop eval and experiment tooling used by researchers.
• Design fast inference paths for model iterations.
• Automate repetitive tasks and improve observability.

Skills

ML training

Distributed computing

Data engineering

Ownership mindset

Tools

Kubernetes

Slurm

Spark

Ray

NCCL basics

Job description

About EkaCare and the mission

EkaCare is India's connected healthcare platform: an EMR that doctors run their practices on, a personal health record used by millions of Indians, and one of the deepest integrations with India's ABDM digital-health rails. Our Parrotlet family of medical models already serves Indian doctors in production, and we open-source our work where it counts.

The role

The versatile builder, you go where the run needs you; data, training, evals, serving, and you leave every system better-instrumented than you found it.

What you'll do

• Build corpus pipelines that feed GPUs without stalls: filtering, dedup, PII redaction, tokenisation, streaming dataloaders at ~500B-token scale.
• Run distributed training jobs on multi-node H200 clusters; make failures loud, recoveries fast, dashboards honest.
• Build the eval and experiment tooling researchers live in.
• Wire fast inference paths (vLLM-class) for model iterations.
• Automate whatever annoyed you last week.

What we look for

• 3–6 years of strong software engineering with real ML training exposure; you've fine-tuned or trained ≥1B-parameter models, not just called APIs.
• Distributed-computing fundamentals: Slurm or Kubernetes, NCCL basics, storage/throughput intuition.
• Data engineering at scale; Spark, Ray, or whatever gets the job done.
• A bias for ownership: you don't ask whose job it is.

Bonus

• Open-source projects or contributions.

Why this is a rare gig

• Open source, with your name on it : weights and technical reports ship publicly.
• India-scale mission : models for a billion people in their own languages.
• Compute that’s rare to fine : dedicated multi-node H200 training under a national grant.
• Small senior team: you work with the people who own the recipe.
• A live deployment path : Government institutes, EkaCare's doctors and patients use what you ship.

Full-Time Employee Benefits

• Medical Insurance & Accidental Insurance