Scale AI · 1 week ago
AI Architect
Scale AI is a company focused on developing reliable AI systems for significant decisions. They are seeking an AI Architect to bridge the gap between AI research and product implementation, driving end-to-end delivery while collaborating with various teams to translate cutting-edge research into impactful solutions.
AI InfrastructureArtificial Intelligence (AI)Data Collection and LabelingGenerative AIImage RecognitionMachine Learning
Responsibilities
Translate research → product: work with client side researchers on post-training, evals, safety/alignment and build the primitives, data, and tooling they need
Partner deeply with core customers and frontier labs: work hands-on with leading AI teams and frontier research labs to tackle hard, open-ended technical problems related to frontier model improvement, performance, and deployment. Shape and propose model improvement work: translate customer and research objectives into clear, technically rigorous proposals—scoping post-training, evaluation, and safety work into well-defined statements of work and execution plans
Translate research into production impact: collaborate with customer-side researchers on post-training, evaluations, and alignment, and help design the data, primitives, and tooling required to improve frontier models in practice
Own the end-to-end lifecycle: lead discovery, write crisp PRDs and technical specs, prioritize trade-offs, run experiments, ship initial solutions, and scale successful pilots into durable, repeatable offerings
Lead complex, high-stakes engagements: independently run technical working sessions with senior customer stakeholders; define success metrics; surface risks early; and drive programs to measurable outcomes
Partner across Scale: collaborate closely with research (agents, browser/SWE agents), platform, operations, security, and finance to deliver reliable, production-grade results for demanding customers
Build evaluation rigor at the frontier: design and stand up robust evaluation frameworks (e.g., RLVR, benchmarks), close the loop with data quality and feedback, and share learnings that elevate technical execution across accounts
Qualification
Required
Deep technical background in applied AI/ML: 5–10+ years in research, engineering, solutions engineering, or technical product roles working on LLMs or multimodal systems, ideally in high-stakes, customer-facing environments
Hands-on experience with model improvement workflows: demonstrated experience with post-training techniques, evaluation design, benchmarking, and model quality iteration
Ability to work on hard, ambiguous technical problems: proven track record of partnering directly with advanced customers or research teams to scope, reason through, and execute on deep technical challenges involving frontier models
Strong technical fluency: you can read papers, interrogate metrics, write or review complex Python/SQL for analysis, and reason about model-data trade-offs
Executive presence with world-class researchers and enterprise leaders; excellent writing and storytelling
Bias to action: you ship, learn, and iterate
Benefits
Comprehensive health, dental and vision coverage
Retirement benefits
A learning and development stipend
Generous PTO
A commuter stipend
Company
Scale AI
Scale’s mission is to develop reliable AI systems for the world’s most important decisions.
H1B Sponsorship
Scale AI has a track record of offering H1B sponsorships. Please note that this does not
guarantee sponsorship for this specific role. Below presents additional info for your
reference. (Data Powered by US Department of Labor)
Distribution of Different Job Fields Receiving Sponsorship
Represents job field similar to this job
Trends of Total Sponsorships
2025 (82)
2024 (54)
2023 (29)
2022 (17)
2021 (10)
2020 (10)
Funding
Current Stage
Late StageTotal Funding
$15.9BKey Investors
MetaAccelTiger Global Management
2025-06-10Corporate Round· $14.3B
2025-06-04Series Unknown
2024-05-21Series F· $1B
Recent News
2026-01-16
CB Insights
2026-01-09
Crunchbase News
2026-01-07
Company data provided by crunchbase