Type
Full-time
The Mission
Join us at the frontier of memory and context engineering for Large Language Models. We're building the governed memory fabric that lets products learn once from truth sources and answer from memory first, turning clever demos into fast, reliable, high ROI AI in the real world. This is category creation: define the primitives, patterns, and proofs that make AI trustworthy and economical at scale.
What You'll Do
- Define the backbone: Establish the core concepts and contracts of an AI memory system, how knowledge is captured, governed, retrieved, and explained.
- Make answers reliable: Design approaches so LLMs find, choose, and use the right context with clear attribution and policy compliance.
- Shape developer & operator experience: Turn complex workflows into simple APIs, UX, and rituals that teams love and adopt.
- Prove value with customers: Partner with design-partners to demonstrate faster responses, lower cost, and stable behavior, then generalize the wins.
- Institutionalize quality: Create lightweight evaluation and observability that quantify accuracy, latency, cost, and trust.
- Build for evolution: Choose abstractions that survive change with versioned contracts, safe rollouts, pragmatic governance.
What You'll Bring
- Product sense for AI systems: You've shipped AI/search experiences that moved a KPI, not just prototypes.
- LLM + IR instincts: Retrieval/ranking, prompting/tool use, and a feel for "useful vs. clever," with measurement to match.
- Systems thinking: You design primitives that are clear today and extensible tomorrow.
- Developer empathy: Clean interfaces, crisp docs, and low-friction onboarding.
- Bias for outcomes: Fast iteration, instrumented learning, evidence-driven scope.
How We'll Measure Success
- •Customer impact: Quantified gains in speed, cost, and answer quality on agreed use cases.
- •Trust & explainability: Every answer carries receipts with source lineage and policy context.
- •Adoption: Time to value quickly; partners expand usage on their own.
- •Resilience: Sensible defaults, safe failure modes, predictable behavior across releases.
How We Work
- •Small surface, high standards: Ship thin, end-to-end slices; refine with live usage.
- •Close to customers: Co-design with demanding partners to reach product truth fast.
- •Governance by design: Safety, policy, and auditability are part of the core, not add-ons.
Compensation & Benefits
We practice pay transparency; ranges reflect US benchmarks and vary by location/experience.
Compensation: Competitive Base + Generous equity
Benefits:
- •Health: Medical, dental, vision
- •Retirement: 401(k) with company match
- •Time off: Flexible PTO
If you're outside the US, we localize benefits and align comp to market while preserving total value.
Why This Role
- •Work on the frontier of LLM memory and context engineering, defining a new layer of AI infrastructure.
- •Own problems end-to-end, from abstraction to adoption and ROI.
- •Build with a small, senior team that values clarity, pace, and integrity.