Senior Applied ML Engineer — LLM Routing (Remote, Equity)

Kimchi; Polska; Remote; PLN 260,000 - 420,000; Full time

Benefits offered by this job

Competitive salary

Remote-first Global environment

Equity options

Learning budget

Job summary

Kimchi is hiring for an applied ML Platform Engineer to own the routing model end to end, including features, objective, training, calibration, and release cadence. You will design production-parity benchmarks, compare routers, and turn fleet data into reliable training signals with careful labeling and error categorization.

You will work on speed-aware routing, calibrated decisions, and abstention under uncertainty, collaborating with a global, remote-first team.

Qualifications

• 5+ years in applied ML or ML research with real-time decision models.
• Strong Python and SQL at scale; can interrogate millions of request-level records.
• Experience with model evaluation and moving from offline to online deployment.
• Prior work on LLM routing and benchmarking is a plus.
• Ability to instrument before optimizing with error bars.

Responsibilities

• Own the routing model end to end: features, objective, training, calibration, and release cadence.
• Build production-parity evaluation and router-vs-router comparisons.
• Turn fleet data into training signals with proper labeling and failure categorization.
• Model the objective the user experiences: task success × token cost × wall-clock speed.
• Run online loops: shadow, A/B, full rollout, drift monitoring, retrain.

Skills

Applied ML

Python

SQL at scale

Model routing

A/B testing

Tools

ClickHouse

PyTorch

GitLab CI

OpenAI API

Job description

Kimchi is hiring for an applied ML Platform Engineer to own the routing model end to end, including features, objective, training, calibration, and release cadence. You will design production-parity benchmarks, compare routers, and turn fleet data into reliable training signals with careful labeling and error categorization.

You will work on speed-aware routing, calibrated decisions, and abstention under uncertainty, collaborating with a global, remote-first team.