Lyft re-architects localization pipeline with iterative LLM Drafter/Evaluator, cutting translation latency from days to minutes
Lyft's localization relied exclusively on human translation with multi-day turnarounds and costs that scaled linearly with each new language. The Quebec launch required Bill 96 French-first compliance faster than multi-day cycles allowed, and European expansion demanded six new languages simultaneously.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Source string submission
Requesters submit source strings along with context about where the text appears in the UI and the intended tone.
Tools used
SmartlingPydanticClaude HaikuGPT's mini models
Outcome
The iterative LLM pipeline achieves a success rate of over 95% across most languages and reduces translation latency from days to minutes, with 95% of translations requiring no significant changes after linguist review.
What failed first
Traditional Neural Machine Translation providers were fast but failed to preserve Lyft-specific terminology and context. Without deterministic guardrails, LLMs consistently hallucinated variable placeholders — the most common failure mode observed in early deployments.