Sarvam DevTools Engineer – Harness Layer
Sarvam
CTC
50L - 70L
Experience
3 - 6 yrs
Location
BLR
123 applications
Sarvam.ai is building foundational AI for India from India!
- Building India’s sovereign Large Language Models (LLMs) and AI infrastructure.
- Focused on multilingual AI systems for Indian languages and enterprise use cases.
- Empowering developers and enterprises with localized, culturally aligned AI.
You'll be a good fit if you have
- Expertise in building SDKs, CLIs, or IDE extensions that improve developer productivity and adoption.
- Experience creating debugging and observability tools to capture invisible failures and make complex agent systems traceable.
- Proven track record of designing developer-first frameworks that optimize workflows with context-aware suggestions and evaluation feedback.
- Strong empathy for developers and the ability to transform agent chaos into structured, reliable, and delightful development experiences.
- Experience implementing time-travel debugging or similar capabilities that allow developers to replay and inspect complex runs.
- Familiarity with evaluation frameworks for performance metrics and regression testing.
- Ability to design observability and telemetry pipelines that provide deep insights into failures.
- Experience with developer-first SaaS, open-source devtools, or observability tooling.
- Strong background in capturing, visualizing, and debugging failures in complex systems.
- A proven ability to optimize authoring workflows and reduce friction through ergonomic tooling.
Key Responsibilities
- Design and Build Developer Tools including SDKs, CLIs, and IDE extensions to improve productivity and adoption.
- Develop Debugging and Observability Features that make complex agent systems traceable and failures visible.
- Create Developer-First Frameworks with context-aware suggestions, evaluation feedback, and workflow optimizations.
- Implement Advanced Debugging Capabilities such as time-travel debugging and replay mechanisms for complex runs.
- Design and Maintain Evaluation Frameworks to measure performance, detect regressions, and ensure reliability.
- Build Telemetry and Observability Pipelines that provide deep insights into failures and runtime behavior.
- Collaborate with Developers to deeply understand pain points and transform agent workflows into reliable, structured, and enjoyable experiences.
- Contribute to Open-Source and SaaS Devtools by shipping ergonomic, developer-first features that reduce friction.
- Visualize and Debug Complex Failures through intuitive interfaces and tooling.
- Continuously Improve Authoring Workflows to streamline developer experience and accelerate agent development.
In the news
You are pre-screened
Team Round1's take
- Selected by the Government of India under the IndiaAI Mission to develop sovereign large language models.
- Committed to open-sourcing its AI models for wider public and enterprise use.
- Combining deep AI research with large-scale Indian language datasets for culturally relevant intelligence.