Skip to content
DevOps & Cloud

Performance Engineer, Inference

Sarvam AI

  • Bengaluru · On-site
  • Full-time · Mid level

Check my ATS match for this job →

Source: Sarvam AI careers page. You apply on the employer's own website.

About this job

Performance Engineer, Inference

Part of Sarvam's Performance Engineering team. We are hiring two specialized performance roles - Inference (this posting) and Kernels (companion posting). They are a vertical stack: the kernels team authors the µs-level GPU code, and the inference team integrates it into a running serving stack and owns the system-level numbers. If your depth genuinely spans both, apply to either and tell us - but most candidates are strongest in one, and we hire for that depth.

Location: [Bengaluru / Chennai / Hybrid / On-site] · Team: Performance Engineering · Level: Senior

About the team

Sarvam serves multiple model families - small and large LLMs, Mixture-of-Experts, Indic ASR & TTS, streaming models and multimodal models - across a multi-node, multi-tenant fleet of Hoppers and Blackwells. The Performance Engineering team owns the numbers the rest of the company plans against: how fast we serve, how much it costs, and how much we get out of every GPU. This team works at the intersection of the serving runtime, the kernel layer, and the SRE org that keeps the fleet alive.

About the role

You will own Sarvam's production serving path for large distributed models end to end. You should be source-level fluent in at least one of SGLang, vLLM, NVIDIA Dynamo, or TensorRT-LLM - able to read and modify it where stock behavior does not fit our workloads - and you should operate and extend a distributed-serving stack at depth: disaggregated prefill-decode across nodes, distributed KV/cache transfer, and the routing and scheduling that span them. You will integrate artifacts from the model and kernel teams into a running multi-node, multi-tenant stack, and you will build and train your own speculators - draft models, distillation from the target, acceptance-rate tuning against the live serving distribution - rather than only wiring in stock implementations. You will produce, and defend, the latency and throughput numbers the company plans against, and you will spend significant time in cross-team work with architecture co-design, the kernels team, the model team, and SRE.

Your scoreboard: TTFT (p50 / p95 / p99), TPOT, throughput, GPU utilization, and cost per million tokens.

What we're looking for

  • 5+ years in ML systems, with 2+ years on inference serving at production scale. Your record shows concrete outcomes - tokens per day, throughput wins, p99 reductions - rather than "deployed a model."

  • Experience serving 100B+ parameter models in production across multi-node tensor, pipeline, or expert parallelism.

  • Source-level fluency in one of SGLang, vLLM, Dynamo, or TensorRT-LLM - you have modified the scheduler, the KV allocator, or the disaggregation path - and reading-level familiarity with the other three.

  • Distributed serving at operating-and-extending depth: a disaggregated prefill-decode stack, distributed KV/cache transfer (Mooncake or equivalent), and cross-node routing and scheduling. You have run one of these in production and modified it where it didn't fit.

  • Speculative decoding as a build-and-train competency: you have trained your own draft models or speculators (EAGLE / DFlash or otherwise), distilled them from a target model, measured and tuned acceptance rate against a real serving distribution, and composed speculation with the rest of the stack - not only integrated a stock implementation.

  • Deep understanding of KV cache internals: block tables, copy-on-write, prefix sharing, and fragmentation.

  • Working command of TP / PP / EP, NCCL primitives, and how they interact with the scheduler.

  • Multi-tenant serving: model co-location and MIG / MPS isolation.

  • Profiling fluency with Nsight Systems, framework tracing, and py-spy / perf.

  • C++ and CUDA at a read-and-modify level.

  • On-call ownership of an inference SLO.

Strong pluses

  • Upstream contributions to SGLang, vLLM, Dynamo, llm-d, TensorRT-LLM, or LMDeploy on non-trivial code paths.

  • Direct production experience with Dynamo or llm-d at scale.

  • Having operated a forked runtime in production.

  • Published or shipped speculator work - a draft model or speculative-decoding technique you trained and measured.

  • MoE serving at scale, long-context (128K+), multi-model serving, or Indic and multilingual workloads.

Skills mentioned

  • LLM
  • C++

Have these skills? Make sure they appear on your resume in the same words.

Similar jobs

See all jobs →
New On-site Senior

Cybersecurity development

Bosch

  • Location:Bengaluru
  • Experience:6+ yrs
  • Employment type:Full-time
  • Salary not disclosed

Job Description JD for Cybersecurity Development: Development, test of HSM and AR Cryptostack Hands-on experience in embedded software development using C/C++, with basic knowledge of scripting languages such as Python for automation and analysis tasks Experience in implementing and supporting au...

  • C++
  • Python
  • Agile
  • Scrum

Bosch careers page

Apply now
New On-site Senior

Sr. Product Owner – DevOps and Software Delivery

Bosch

  • Location:Bengaluru
  • Experience:14–15 yrs
  • Employment type:Full-time
  • Salary not disclosed

Job Description Own the product roadmap, backlog, and release planning for enterprise software delivery platforms.Drive end-to-end software delivery across development, testing, deployment, and operations.Lead stakeholder management across business units, engineering teams, architects, operations...

  • Stakeholder Management
  • CI/CD
  • Agile
  • Scrum
  • Azure
  • Kubernetes
  • +7 more

Bosch careers page

Apply now
On-site Lead / Manager

Staff DevOps Engineer

Zscaler

  • Location:Bengaluru
  • Employment type:Full-time
  • Salary not disclosed

Zscaler (NASDAQ: ZS) accelerates digital transformation so customers can be more agile, efficient, resilient, and secure. The Zscaler Zero Trust Exchange platform protects thousands of customers from cyberattacks and data loss by securely connecting users, devices, and applications in any locatio...

  • AWS
  • GCP
  • Kubernetes
  • Terraform
  • CI/CD
  • Ansible
  • +3 more

Zscaler careers page

Apply now
On-site Entry level

IT System Administrator I

Rubrik

  • Location:Bengaluru
  • Experience:5+ yrs
  • Employment type:Full-time
  • Salary not disclosed

About The Team: Come join our growing team to deliver amazing technology support experiences that fuel the vision of Dynamic Work for our employees. As a System Administrator, you will focus on delivering world-class technology support for both our employees and our workspaces, driven by the impo...

  • SaaS
  • Jira
  • Onboarding
  • Python
  • Confluence

Rubrik careers page

Apply now
On-site Lead / Manager

OT Security Consultant and Cybersecurity Solution Architect

Bosch

  • Location:Bengaluru
  • Experience:10–14 yrs
  • Employment type:Full-time
  • Salary not disclosed

Job Description Roles & Responsibilities : OT Asset Discovery and Validation: Discover OT assets, connected devices and communication protocols across all in-scope plants, including unmanaged assets. Apply a safe mix of passive and active discovery and data collection methods and validate every f...

  • Stakeholder Management

Bosch careers page

Apply now
On-site Lead / Manager

Engineering Manager, Core Infrastructure

Stripe

  • Location:Bengaluru
  • Experience:10+ yrs
  • Employment type:Full-time
  • Salary not disclosed

Who we are About Stripe Stripe is a financial infrastructure platform for businesses. Millions of companies—from the world’s largest enterprises to the most ambitious startups—use Stripe to accept payments, grow their revenue, and accelerate new business opportunities. Our mission is to increase...

  • AWS
  • Kubernetes
  • Ruby
  • Java

Stripe careers page

Apply now