EU Blue Card eligible · open to relocation · available Q3 2026
I'm Saurabh Joshi, a data scientist with five years shipping ML pipelines that run in production — not notebooks. Currently relocating to Germany for a senior ML role.
A binary classifier predicting which healthcare claims need a specific revenue-cycle action, refactored from a notebook POC into a scheduled, versioned production pipeline.
A console app for CARC/RARC weighted claim scoring, rebuilt as a containerized FastAPI service and enterprise proof of concept.
A configurable PySpark service for pulling and deduplicating allocation data across workspaces, with the config-driven query layer that keeps it flexible.
A from-scratch MLOps capstone: a retrieval-augmented agent backed by an MCP server and a LoRA-fine-tuned extraction model, with an evaluation harness.
I build the unglamorous parts of machine learning well: incremental extraction, partitioned storage, retry logic, model versioning, and serving — the layer between a good model and one a team can actually depend on. My home turf is healthcare revenue cycle management, where a misfired prediction has a real cost attached.
I'm honest about where I'm growing. I'm investing deliberately in transformer internals, LLM fine-tuning, and MLOps depth — and the work shows up in my recent projects rather than just on a skills list.
I'm relocating to Germany and looking for a senior data science / ML engineering role in Berlin or Munich.
// status: in development
Interactive, in-browser ML demos are landing here soon — a threshold explorer over a real classifier and a client-side inference demo, both running entirely in your browser. No server required.
updated June 2026
A proof of concept runs once, on your machine, while you watch. Production runs unattended, forever. Here's the refactor that bridged the two for an RCM classifier.
An AUC of 0.90 told me the model was good. A 0.5 cutoff told the business it was useless. Here's how I closed that gap on a production RCM classifier.
I'd like to hear about it. The fastest way to reach me is email.
hello@saurabhjoshi.dev