Clari
About this role
Clari and Salesloft have merged and now build what they call a predictive revenue system, software that helps companies forecast and manage sales. This role is on the agentic AI side and the posting is unusually blunt about what it wants: a builder, not a manager in waiting, and an execution role rather than a research one. The work is taking an ambiguous problem, such as whether an agent can safely update a CRM field, and turning it into a working prototype in days, then into production quality Python. It asks for six to eight years total, split as two to three years shipping LLM inference, RAG or agent work on top of two to three years of core NLP and data science before that. Bengaluru, hybrid.
Who this is for
What the posting asks for:
- 6 to 8 years total experience, made up of 2 to 3 years hands on in agentic AI and 2 to 3 years in core NLP and data science before that.
- Real experience shipping LLM based inference, RAG or agent work, including multi agent systems.
- Core NLP and ML foundations: text classification, named entity recognition, embeddings and feature engineering, from before the LLM era.
- The judgement to know when classical NLP or ML is the better fit and when an LLM approach is, which the posting says should come from having actually done both.
Day to day:
- Prototyping fast: turning a rough problem statement into a working RAG pipeline, LLM inference or agent within days rather than weeks.
- Taking your own prototypes from notebook to production quality Python: tested, typed and reviewable.
- Partnering with the platform engineers who own the harness and infrastructure to ship. You are not expected to design the harness, but you need to understand it well enough to build in it and debug it.
- Owning evaluation for everything you ship: offline eval sets, LLM as judge checks, and tracking regressions before and after changes.
- Pulling in proofs of concept and external research when the path is unclear, and course correcting quickly.
Location and office reality: the posting states this is a hybrid position requiring the ability to be onsite at the Bengaluru office as needed, and that candidates must be based in India.
Honest fit guidance: the posting says outright it is looking for a builder rather than a manager in waiting, and that most of your time goes on prototyping and evaluation. If you want a research role or a team lead track, this is not it. The two to three years of pre LLM NLP experience is a genuine requirement and rules out people who came to the field entirely through prompt engineering.
- 6 to 8 years total experience, made up of 2 to 3 years hands on in agentic AI and 2 to 3 years in core NLP and data science before that.
- Real experience shipping LLM based inference, RAG or agent work, including multi agent systems.
- Core NLP and ML foundations: text classification, named entity recognition, embeddings and feature engineering, from before the LLM era.
- The judgement to know when classical NLP or ML is the better fit and when an LLM approach is, which the posting says should come from having actually done both.
Day to day:
- Prototyping fast: turning a rough problem statement into a working RAG pipeline, LLM inference or agent within days rather than weeks.
- Taking your own prototypes from notebook to production quality Python: tested, typed and reviewable.
- Partnering with the platform engineers who own the harness and infrastructure to ship. You are not expected to design the harness, but you need to understand it well enough to build in it and debug it.
- Owning evaluation for everything you ship: offline eval sets, LLM as judge checks, and tracking regressions before and after changes.
- Pulling in proofs of concept and external research when the path is unclear, and course correcting quickly.
Location and office reality: the posting states this is a hybrid position requiring the ability to be onsite at the Bengaluru office as needed, and that candidates must be based in India.
Honest fit guidance: the posting says outright it is looking for a builder rather than a manager in waiting, and that most of your time goes on prototyping and evaluation. If you want a research role or a team lead track, this is not it. The two to three years of pre LLM NLP experience is a genuine requirement and rules out people who came to the field entirely through prompt engineering.
Apply on company site
Opens www.salesloft.com, the employer's own application page. Applying is always free.