Tech Lead - Atlas (Internal Agent Engineering)

About Basis

Basis equips accountants with a team of AI agents to take on real workflows.

We have hit product-market fit, have more demand than we can meet, and just raised $34m to scale at a speed that meets this moment.

Built in New York City. Read more about Basis here.

About The Team

Atlas is Basis’s internal agent team with a single mandate: make every employee 100x more productive.

We’re building the systems, agents, and automations that make Basis the first truly agent-native company. Atlas operationalizes knowledge, connects context, and builds the tools that let engineers, accountants, and operators move faster and think bigger.

Our work spans everything from internal coding agents and evaluation harnesses to knowledge graphs, retrieval pipelines, and Retool apps. We treat Basis itself as the ultimate testbed for AI-driven organizations—turning every workflow into a feedback loop between people and agents.

About The Role

As a Tech Lead on the Atlas team, you’ll hold the technical vision for how internal agents, automations, and context systems work together. You’ll design and build the architecture that powers Basis’s internal intelligence layer—and teach others how to build within it.

You’ll move fast, prototype constantly, and ship improvements that compound leverage across the company. Your work will span infrastructure, ML integration, and developer experience: connecting code, data, and process through intelligent automation.

You’ll lead by example—writing code, reviewing designs, and defining patterns that make Basis itself a proving ground for how agent-native systems should operate.

What You’Ll Be Doing:

1. Architect the agent and context systems that power Basis

• Design internal agent architectures for coding, operations, and research workflows.
• Define standards for prompts, tools, and context management; make information architectures agent-friendly.
• Build APIs and services that let agents access the right data safely and efficiently.
• Create frameworks for evaluating agent behavior—success criteria, trajectory analysis, and continuous improvement loops.

2. Build high-leverage internal tools and automations

• Ship Retool apps, Chrome extensions, Slack bots, or MCP integrations that automate repetitive workflows.
• Identify bottlenecks across engineering, accounting, and ops; prototype and ship fixes quickly.
• Integrate SaaS systems, repos, and APIs into coherent automation surfaces.
• Balance no-code/low-code flexibility with robust engineering when complexity warrants it.

3. Design the foundation for agent-native knowledge and retrieval

• Map Basis’s internal knowledge graph—docs, data, and code—and make it navigable to both humans and agents.
• Implement retrieval and embedding pipelines to power context-aware automation.
• Define the schemas and interfaces that keep context stable and versioned over time.
• Continuously improve how agents learn from Basis’s evolving internal state.

4. Lead through clarity and experimentation

• Hold the architectural vision for internal agents and context systems.
• Run sharp design reviews and retros; document trade-offs and decisions clearly.
• Mentor engineers through pairing, design discussions, and post-mortems.
• Model velocity with discipline: ship fast, measure impact, and refine relentlessly.

📍 Location: NYC, Flatiron office. In-person team.

What Success Looks Like In This Role

• Architect: You’ve built an internal agent or automation system that others depend on daily.
• Integrator: Knowledge, data, and tooling connect cleanly through your abstractions.
• Force multiplier: Every week, something you built makes the company faster or smarter.
• Teacher: Engineers and operators around you learn to reason clearly about agents, context, and automation.
• Builder: You approach every project with clarity, conviction, and curiosity.
• Architect: You’ve built an internal agent or automation system that others depend on daily.
• Integrator: Knowledge, data, and tooling connect cleanly through your abstractions.
• Force multiplier: Every week, something you built makes the company faster or smarter.
• Teacher: Engineers and operators around you learn to reason clearly about agents, context, and automation.
• Builder: You approach every project with clarity, conviction, and curiosity.

In accordance with New York State regulations, the salary range for this position is $100,000 –$300,000. This range represents our broad compensation philosophy and covers various responsibility and experience levels. Additionally, all employees are eligible to participate in our equity plan and benefits program. We are committed to meritocratic and competitive compensation.