LinkedIn and ATS tools count keywords. Ganloss checks what candidates actually shipped — repos, papers, prod, tests.
Recruiter guide · matching on shipped projects
Remote Discord LLM engineer for community copilots and summarization — product GenAI with strong safety bars.
Remote US Groq ML systems role — runtime, scheduling, and developer APIs for high-throughput LLM inference.
Groq LLM inference engineer jobs — ultra-low-latency serving on LPUs, compiler tooling, and API performance in Mountain View.
ServiceNow MLOps engineer for model CI/CD and GPU serving — Bellevue / Seattle hybrid enterprise AI platform.
ServiceNow AI jobs: LLM engineer on Now Assist for IT and HR workflows — enterprise RAG, guardrails, and evals in Santa Clara.
Dropbox LLM engineer for AI writing and file copilots — Seattle hybrid with TypeScript + Python stack.
Dropbox Dash RAG engineer — enterprise knowledge retrieval, citations, and permission-aware search in San Francisco.
Remote US Pinterest GenAI role for generative Pin experiences — diffusion, evals, and creator-facing APIs.
Pinterest ML engineer jobs on visual search and embedding retrieval — computer vision for Pins in San Francisco.
DoorDash GenAI for merchant menus, support, and ops copilots — Seattle hybrid LLM engineering.
LinkedIn AI jobs: LLM engineer on Job Matching and Recruiter copilots — retrieval, ranking, and fairness evals in Sunnyvale.
Uber GenAI engineer building rider and driver support agents with RAG, tool use, and eval harnesses — Seattle hybrid.
An “LLM engineer” profile can hide zero lines of production RAG code. Recruiters lose weeks sorting identical ChatGPT-inflated résumés.
Ganloss flips the logic: matching starts from shipped projects (GitHub, Kaggle, papers, technical cases) before the first interview.
1) Define expected proof (repo, deployment, metric). 2) Source on Ganloss + filter by stack. 3) Short practical case (2–3 h). 4) Close with a transparent package.
Result: fewer false positives, shorter timelines, better post-hire retention.