[VMT] Platform AI Software Engineer
Software Mind3 days ago
Buenos Aires, ArgentinaSenior
Responsibilities
- Investigate agent misbehavior from customer plants and turn each case into a diagnosis, fix, and regression test.
- Build and extend evaluation harnesses that gate prompt changes and model migrations.
- Develop reliability mechanisms including assumption surfacing, grounding checks, and output-quality reviewers.
- Diagnose and resolve cross-cutting performance and concurrency issues.
- Improve code quality in touched areas while following team review conventions.
Requirements
- Strong Python experience in large, shared, evolving backend codebases.
- Experience shipping an LLM-based feature to production and building an evaluation that changed a real product or model decision.
- Experience debugging production issues from symptoms through confirmed root causes, including latency spikes, concurrency errors, and silent failures.
- Experience treating prompt work as engineering through measured adherence and regression testing against fixed case sets.
- Comfort with feature flags, staged rollouts, small pull requests, and mostly asynchronous collaboration across US and Australia time zones.
- High autonomy in scoping and delivering fixes for ambiguous problems, along with direct and precise written English.
- Preferred: GCP, especially Vertex AI; AWS or Azure; observability tooling; charting or plotting pipelines such as matplotlib; industrial, process, or time-series data experience; and AI-tooling workflows such as Claude Code.
Benefits
- Educational resources.
- Flexible schedule and work-from-anywhere arrangement.
- Referral Program.
- Supportive and chill atmosphere.
- Applications accepted from LATAM countries.
Tech Stack
Categories