Loading...
Product Lead, Safety Systems & Trust at Inflectionai
InflectionaiPalo Alto, California, United States·1d ago
LeadFull-timeRustRailsAIML
Salary Range
$165,750 - $224,250 /year
This salary is estimated based on similar roles. The actual salary may vary.
About the Role
Inflection AI creates high-EQ AI agents that people and brands can trust. As creators of the original high-EQ frontier models, we continue to view Trust & Safety as foundational as we develop agents that are emotionally intelligent, deeply capable, and radically steerable.
Key Responsibilities
- →Productize safety stack and inference engine capabilities, including low-latency, token-level safety filters and constrained decoding protocols that block harm without degrading performance.
- →Partner with Research on alignment and model behavior, defining RLHF/DPO objectives and reinforcement signals while developing taxonomies for steerability and model behavior.
- →Lead adversarial assessment and red-teaming initiatives, architecting automated stress-test infrastructure and evaluation frameworks for safety benchmarks.
- →Serve as a strategic leader driving cross-functional teams to implement and evolve Inflection’s most critical safety initiatives.
- →Build the affordance layer for trust, working with Design to ensure users understand when and why to trust AI interactions and agent decisions.
- →Drive policy-as-code execution, partnering with Legal and Engineering to translate privacy, safety, and brand principles into technical specifications.
- →Shape system instructions, fine-tuning datasets, and user experience guardrails to support safe and ethical AI deployment.
Requirements
- ✓Technical deep-diver with a strong grasp of ML concepts including LLMs, RAG, diffusion models, and the nuances of model drift.
- ✓8+ years of product experience, with a track record of shipping complex products from 0 to 1 at scale, ideally in Integrity, Trust & Safety, or other high-stakes AI domains.
- ✓Systems thinker who thrives in high-ambiguity, low-precedent environments where you define both the problem and the solution.
- ✓Ethical architect and first-principles thinker committed to making AI safe and beneficial for humanity.
- ✓Hands-on experience with Constitutional AI or self-correction loops in LLM chains.
- ✓Deep expertise in adversarial analysis, including many-shot jailbreaking and prompt injection mitigation.
- ✓Proficiency in creating safety evaluations and using telemetry to measure system performance and blind spots.
- ✓Have a bachelor’s degree or equivalent in a related field to the offered position requirements.
