Staff AI Platform Engineer: Build Next-Gen Agentic AI

Sentry; San Francisco (CA); In loco; USD 240.000 - 400.000; Tempo pieno

Vantaggi offerti da questo lavoro

Commuter Stipend

Professional Development

Flexible PTO

Charitable Matching

Yoga breaks

Company match program

Descrizione del lavoro

Sentry is seeking a Staff engineer to evolve its AI debugger, Seer, and to decide how AI systems should grow as models evolve. You will lead end-to-end AI initiatives, work with developers and product teams, and deploy in production while continuously debugging and improving performance.

You will shape how agents obtain context, use tools, and determine completion, building evaluation frameworks to validate improvements fast and cost-effective.

Competenze

• Shipped an LLM-backed system and can explain what broke.
• Strong opinions on agent design and have changed at least one in the last year.
• Trained or fine-tuned models when appropriate and know when not to.
• Proficient in production Python.
• Led work across multiple engineers and teams with design docs that are read.

Mansioni

• Build state-of-the-art agentic AI systems to triage, debug, and solve real production issues.
• Set technical direction for how agents get context, use tools, and decide when they're done.
• Build evals to measure model/prompts impact quickly for adoption.
• Decide when to use frontier vs smaller/trained models as cost/capability shift.
• Leverage Sentry’s dataset of errors, spans, and profiles.
• Own major AI initiatives end to end with developers and product teams.

Conoscenze

LLM-backed systems

Agent design

Production Python

Cross-team leadership

Technical documentation

Descrizione del lavoro

Sentry is seeking a Staff engineer to evolve its AI debugger, Seer, and to decide how AI systems should grow as models evolve. You will lead end-to-end AI initiatives, work with developers and product teams, and deploy in production while continuously debugging and improving performance.

You will shape how agents obtain context, use tools, and determine completion, building evaluation frameworks to validate improvements fast and cost-effective.