Member of the Technical Staff

Poetiq
United States
Workplace: OnsiteFull timeUSD 150,000 - 500,000 annuallyFunction: Software EngineeringSkills: ["Autonomy","Communication","Rigorous testing","Architectural design","Performance profiling"]

Build and improve recursive self-improvement (RSI) systems by designing core algorithms and pushing novel reasoning strategies. You’ll create dynamic execution harnesses, architect orchestration for highly parallel LLM workflows, and build diagnostic pipelines to pinpoint knowledge gaps vs. tool and reasoning failures. Close the feedback loop by routing failures into SFT data pipelines and RL tasks, while running reliable high-performance infrastructure against third-party LLMs—all fully on-site in Los Altos.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Poetiq
Poetiq
20 hours ago

Member of the Technical Staff

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 1 hour agoStatus: Live

Job Summary

Build and improve recursive self-improvement (RSI) systems by designing core algorithms and pushing novel reasoning strategies. You’ll create dynamic execution harnesses, architect orchestration for highly parallel LLM workflows, and build diagnostic pipelines to pinpoint knowledge gaps vs. tool and reasoning failures. Close the feedback loop by routing failures into SFT data pipelines and RL tasks, while running reliable high-performance infrastructure against third-party LLMs—all fully on-site in Los Altos.
Location: United States
Workplace: Onsite
Employment Type: Full time
Job Function: Software Engineering
Seniority: Sr. Manager level

Key Responsibilities

  • •Design and build core algorithms that drive system-level self-improvement in the RSI loop.
  • •Propose and test novel, task-specific reasoning strategies; probe LLMs to surface latent knowledge; build data-efficient methods to find optimal strategies.
  • •Synthesize dynamic execution environments on the fly and architect orchestration for highly parallel, asynchronous LLM interactions.
  • •Build diagnostic pipelines to separate genuine knowledge gaps from tooling limits and reasoning/composition failures.
  • •Route knowledge-gap failures into SFT data pipelines and convert reasoning/composition failures into verifiable, exact-match RL tasks; run infrastructure at scale for external LLMs.

Pay and Benefits

Salary: USD 150,000 - 500,000 annually
Equity and Bonus:Equity
Perks:EquityAnnual Bonus

Key Requirements

  • •Autonomy to turn well-defined and highly ambiguous ideas into production-grade reality.
  • •First-principles understanding of LLM reasoning and failure modes, with hands-on experience with LLM “quirks.”
  • •Ability to build loops where the system learns from each solved problem and improves on the next.
  • •Commitment to code quality, rigorous testing, performance profiling, deliberate architectural design, and clear communication.
  • •Fully on-site in Los Altos (in-person team; a fit requirement).
Experience:AI research
Skills:AutonomyCommunicationRigorous testingArchitectural designPerformance profiling
Tech Stack:LLMsSFTReinforcement learning (RL)Data pipelinesDiagnostic pipelinesHigh-performance infrastructure

Company Brief

Poetiq
Develops AI-driven creative writing tools that generate poetry and literary content, enabling users and brands to produce personalized, stylistic poems and short-form creative text using machine learning models.
Industry: Digital Media
Website