Member of Technical Staff - Atlas at Basis, New York

Basis builds real agents that perform end-to-end work for some of the largest accounting firms in the world. Now, its internal Atlas team is tackling a deeper question: what does knowledge work look like once we reach AGI? The Member of Technical Staff - Atlas role in New York, NY is for engineers who want to spend the next several years making that answer real.

Building the First AI-Native Company at Scale

Basis recently raised $100M at a valuation exceeding $1B, with backing from Khosla Ventures, Google Ventures, Accel, and notable operators and technical leaders including Jeff Dean, Nat Friedman, Daniel Gross, and Adam D'Angelo. OpenAI's Applied AI Lead, Prashant Mital, noted that Basis has "pushed the limits of what we thought our models could do" on complex, economically valuable accounting tasks.

Atlas began when Basis had fewer than forty people. Since then, the company has secured major funding, giving Atlas the resources, access, and mandate to pursue its mission seriously. The team embeds directly within Basis's operations to build intelligence deeper into workflows.

Atlas has already shipped:

• Agent teammates that push code directly to production
• Systems that keep company knowledge current automatically
• Services that diagnose and repair themselves

The ambition goes far beyond making someone 10% more productive. Atlas is building zero-human loops that handle entire jobs end-to-end, freeing the team to take on higher-level work. Getting there means inventing new infrastructure, interfaces, operating models, and ways for humans and agents to collaborate. There is no established playbook for that—building it is the job.

What Kind of Engineer Thrives Here

Basis hires individuals, not resumes. Given how fast the landscape is changing, finding the best people matters more than matching specific experience.

First Principles and Systems Thinking

Many of the problems Atlas solves are net new. The team wants engineers who can decompose novel problems, reason from fundamentals rather than pattern match, and design systems that compound.

Thinking Beyond the Engineering

The strongest engineers here understand the domain deeply—the data systems, workflows, and tasks agents perform. That understanding is what enables better system design, better agent training, and sharper product decisions.

Ownership of Outcomes

Every project at Basis has a single Responsible Party (RP) who is accountable for whether it ships and whether it works. There are always more things to own than people to own them. Engineers own their work end-to-end.

Agency

Engineers need the initiative to gather context, talk to the right people, read data, and understand problems deeply enough to know what matters.

Non-Negotiables

• In-person at the Flatiron office in NYC five days a week
• Excitement about working with coding agents—every engineer gets an unlimited token budget
• Passion for applying agents to real-economy problems at the frontier

Benefits Package

Basis offers a thoughtful benefits program supporting physical, mental, and financial well-being:

• Health & Wellness: Premium medical, dental, and vision coverage; life insurance; six coaching and six therapy sessions through Spring Health; 100% covered Wellhub Gold membership (includes gyms like Lifetime, SoulCycle, NYSC plus apps like Strava and Headspace)
• Time Off: Unlimited PTO and 12 paid company holidays
• In-Office Perks: Daily meal stipends, fully stocked kitchen, $300 toward custom desk setup
• Financial: Pre-tax commuter benefits and 401(k) retirement plan
• Culture: Monthly office activities and optional team happy hours
• Parental Leave

Application Tips for This Role

Since this position involves building production-grade autonomous agents at scale, you'll want to showcase competencies that align with Basis's engineering philosophy:

• Demonstrate systems-level thinking: Highlight projects where you designed infrastructure from scratch rather than extending existing patterns. Atlas needs engineers who decompose novel problems from first principles.
• Show agent-building experience: If you've worked with LLM frameworks or built tool-calling pipelines, make that prominent. Even personal projects involving multi-step agentic workflows demonstrate relevant capability.
• Highlight domain depth: Basis values engineers who understand the domain their systems serve. If you've worked in accounting tech, data pipelines, or workflow automation, connect that experience to what Basis's agents do for accounting firms.
• Reference your RP mindset: Mention instances where you owned a project end-to-end and shipped it. Basis's Responsible Party system means accountability for outcomes, not just output.
• Engage with the AI agent community: Follow discussions on platforms like LessWrong, the Latent Space podcast, and Twitter/X where frontier agent research is shared. Mentioning specific advances in function calling or multi-agent orchestration shows genuine engagement with the field.
• Prepare a technical portfolio: Given the unlimited coding agent token budget, showing how you've leveraged tools like Cursor, Copilot, or Cline to accelerate complex engineering work signals readiness for their workflow.