What will be on job:

Work with real customer data - error analysis, understanding problems etc.

Develop algorithms

Synthetic data generation

RAG

Fine tuning

Agents

Inference optimization

Write production code

Algorithms in a scalable manner

Data pipelines

Evaluation infrastructure

Must have - 2 types of profiles:

From industry

Passion and knowledge in the domain - up to date with models, techniques etc.

Background with production systems - as a software or LLM engineer - at least 4 years in total

If software background only - needs to have experience with data

Academic

Passion and knowledge in the domain - up to date with models, techniques etc.

At least masters degree - in NLP / LLMs

At least 2 years as a software engineer / algorithms engineer working on production systems in a data environment

Advantage:

Built production LLM application in the past

Past in startup companies