AI Native Builder/Architect

About the Role

Atlan is building the context layer for enterprise AI, the essential infrastructure that AI agents require to understand business context around data. Gartner has identified context as the defining issue in enterprise AI and context graphs as essential infrastructure for AI agents. Atlan is recognized as a leader in data catalogs, data governance, and metadata management, the foundational elements for this context layer.

The Team

You will join the team responsible for building Atlan's context layer, the enabling platforms, and vertical agents that operate on it. This platform provides agents with necessary components like a registry of existing agents and skills, traces and evaluations for every run, and governance over cost and access. Vertical agents perform specific jobs for customers and internal teams, with existing agents already managing real work within Atlan. The team operates in small "two-slice" units, where individuals own and build initiatives with one or two collaborators.

Key Responsibilities

• Own 0→1 initiatives from problem definition to production, including framing, specification, architecture selection, and implementation.
• Build agents collaboratively with stakeholders, scoping workflows, demoing early, and deploying agents within existing operational workflows (e.g., Slack approvals, Word document redlines).
• Earn autonomy by starting with shadow runs, analyzing errors, and expanding agent capabilities as results improve.
• Design for failure by defining behavior when tools are unavailable, context is limited, or the agent is uncertain, and establishing human handover processes.
• Make engineering decisions regarding the balance between deterministic code and models, build vs. reuse, model selection, cost, and reliability, clearly explaining trade-offs.
• Ensure agent behavior is observable by linking traces and evaluations to specific revisions, tracking consistency, cost, and latency alongside success rates, and treating drift as a regression.
• Build upon and extend the context layer, agent platform, and registry by implementing versioned skills, layered instructions, explicit permissions for tools and secrets, and managing model changes as migrations with pre-promotion evaluations.
• Build AI-native systems by utilizing coding agents for personal work, maintaining their skills and instructions, and reviewing their output for architectural consistency.
• Share learnings by converting effective practices into adoptable skills, reference architectures, and quality gates, and demoing developed solutions.

Requirements & Tech Stack

• Hands-on experience building software systems and shipping 0→1 products or platforms into production, often without a formal spec.
• Experience shipping agents or LLM systems used by real individuals.
• Proficiency in tracing issues in agents or LLM systems back to the root cause (context, tool call, or decision).
• Experience using evals to identify and resolve problems, validate scores, and judge agents by user impact (hours saved, error rate, cycle time).
• Demonstrated ability to design for failure, considering potential agent actions, consequences of errors, and human approval points.
• Experience using coding agents daily and holding their output to high standards.
• Critical optimism: belief in agent capabilities combined with rigorous testing of failure points.
• Ability to reason from first principles and rebuild systems when familiar designs or models become outdated.
• Strong systems judgment for designing and building production-ready systems with attention to reliability, latency, cost, and security, and ensuring they are extensible by others.
• Effectiveness in owning initiatives with one or two collaborators, without requiring a large team.

Strong Bonus

• Experience with MCP, A2A, agent frameworks and runtimes, LLM gateways, agent observability, or building an agent harness or orchestration layer.
• Experience with knowledge graphs, vector search, retrieval, or semantic layers.
• Experience building an eval suite from scratch, or conducting shadow evaluations, LLM grading, or staged rollouts.
• Experience shipping agents that write to a system of record with permissions, approvals, and an audit trail.
• Experience leading AI adoption within an engineering or business team, with demonstrable before-and-after impact.
• Founding-engineer, technical-founder, or early-stage startup experience, involving building solutions from inception.
• A public record of work, including repos, demos, write-ups, or talks, showcasing your building process.

Why Atlan?

• Competitive Compensation: Benchmarked at the top of the market, with strong base salary, performance-based variable pay, and impact-driven equity.
• AI Native Culture: Build the future of work systems with AI integrated into how we build, think, and work.
• Health & Wellness: Comprehensive benefits including health, dental, vision, mental health, and flexible stipends.
• Flexible Time Off & Leave Policies: Trust-based policies to ensure proper rest and support.
• Accelerated Growth & Learning: Develop through cutting-edge technology, complex implementations, and an experienced team.
• Global, Remote-First, High-Trust: Work from anywhere in a flexible, asynchronous, and trust-first environment.