The Moment

Search Atlas bootstrapped to $35M ARR by solving a problem most companies don't even attempt: autonomous marketing execution at enterprise scale.

We've cracked the code on daily autonomous deployment. We've proven that AI can generate, review, and ship production code reliably. We've scaled agentic systems to orchestrate hundreds of integrated services.

Now we're hitting the walls that only frontier companies face. The problems we're solving, perpetual agent coordination, cost-quality tradeoffs in LLM orchestration, distributed state management in autonomous systems, aren't taught in universities or discussed in most engineering organizations.

You're the person who's stared into these kinds of problems before.

We're not hiring a traditional VP. We're looking for a engineer who leads, in the mold of Staff+ ICs at Anthropic, OpenAI, or the infrastructure teams at hyperscalers.

What You'll Actually Do:

Daily Technical Reality (60% of your time)

• Whiteboard architecture for systems running autonomous agents continuously, where every architectural decision impacts token costs, execution latency, and quality;

• Review critical-path code in our agentic orchestration layer, not to micromanage, but because decisions here ripple to millions of requests;

• Solve distributed systems problems that arise when coordinating multiple LLM-based agents with bounded autonomy, managing context windows, and optimizing state persistence;

• Unblock engineers by doing hands-on technical work: refactoring a bottleneck, designing a new subsystem, debugging a production incident that requires deep systems thinking

• Make real-time tradeoff decisions: Cost vs. quality. Speed vs. reliability. Cheaper models vs. frontier models. You'll spend time analyzing token efficiency, model performance, and execution patterns.

• Code daily. Not ceremonial code reviews or architecture docs. Real code. You ship features, you optimize critical paths, you own the forensics when something breaks.

Leadership & Scaling Reality (40% of your time)

• Lead 100+ engineers through hands-on mentorship, not management theater. Your team gets better by watching you solve a hard problem, not by sitting through standup meetings.

• Identify and develop the next generation of technical leaders. You know how to spot an operator in 15 minutes and distinguish them from someone who looks good in meetings.

• Set technical standards that elevate the entire organization. TDD, trunk-based development, aggressive deployment cadence, not as buzzwords, but as lived practice.

• Shape engineering culture around ownership: end-to-end responsibility, no project managers, no blame games. Engineers own what they build, all the way to production and monitoring.

The Technical Context

The Complexity You'll Face

Multi-Agent Orchestration at Scale:

• Coordinating autonomous agents across dozens of decision points

• Managing context window constraints while maintaining quality

• Designing agent architecture where each component must be reliable (failure in one agent cascades)

• Balancing tool availability vs. token efficiency vs. execution speed

LLM-Driven Code Generation in Production:

• ~70% of our production code is generated by agentic systems

• Continuous deployment of AI-generated changes (daily releases)

• Evaluation frameworks that catch quality degradation before it reaches production

• Prompt versioning, golden datasets, and automated testing for model changes

• Understanding when to use different models for different tasks, optimizing for cost and quality

Long-Running Autonomous Systems:

• Agents that need to operate perpetually, learning from each execution

• State management across distributed agents without degradation

• Memory constraints in systems designed to run for days/weeks

• Error recovery and circuit breakers when things inevitably break

Cost-Quality-Speed Trifecta:

• Managing token costs at scale (millions of LLM calls daily)

• Quality degradation when moving away from frontier models

• Latency requirements (sub-100ms critical path)

• Architectural decisions that compound costs: choosing the wrong model, over-contextualizing agents, or poor tool management can cost millions monthly

Integration Complexity:

• Orchestrating dozens of microservices as "tools" available to agents

• Handling microservice reliability, failures, and inconsistent responses

• Designing APIs that agents can reliably interact with

• Managing the blast radius of changes across tightly coupled systems

The Stack You'll Inherit

Core Infrastructure:

• Backend: Python (Django, FastAPI), async processing at massive scale (Celery/RQ)

• Data Layer: PostgreSQL (transactional), ClickHouse (analytics), Elasticsearch, Redis

• Compute: Kubernetes, Docker, multi-region coordination

• Monitoring: Datadog, Sentry, Grafana (real-time visibility into agentic systems)

• AI/ML: Vector databases, LLM APIs (Anthropic, OpenAI, Google), prompt management, evaluation frameworks

What Makes It Hard:

• Not a typical SaaS stack, this is a systems company built for continuous autonomous execution

• Scale you can't think your way out of; you have to engineer solutions

• Tight coupling between business logic, AI systems, and infrastructure (changes in one layer affect all)

• Real-time requirements competing with cost constraints

Who We're Actually Looking For

The Profile

You have 10+ years of shipping production code, with 3+ years of hands-on technical leadership scaling engineering organizations from 30-100+ people. But here's the catch: you didn't stop coding when you became a leader. You're the person who:

• Has operated at frontier: You've worked at Anthropic, OpenAI, a major hyperscaler, or a well-funded startup solving problems at the edge of what's technically possible

• Understands agentic systems (not as theory, but from shipping production code): Multi-agent orchestration, LLM integration, reliability under uncertainty, cost optimization in AI systems

• Has hit scaling walls you had to engineer your way out of: Token management, context window constraints, perpetual system design, distributed state, cost-quality tradeoffs

• Lives and breathes shipping: You measure cycle time, deployment frequency, MTTR. You see waste and immediately think "how do we eliminate this?"

• Attracts elite engineers: The top 1% of SF engineers want to work for you because they know you'll make them better. You have a reputation for technical rigor and mentorship through action.

• Combines depth with breadth: You're comfortable reviewing code at any layer - infrastructure, backend, frontend, ML ops. You know when you're out of your depth and have no ego about learning.

• Thinks like a founder: You understand unit economics, customer impact, and how engineering decisions compound. You're not optimizing for resume-building; you're optimizing for what matters.

What You're Not

• A manager who delegates all code to direct reports

• Someone who measures success in headcount or org charts

• A process engineer ("let's add more meetings to align on decisions")

• Someone uncomfortable with ambiguity or unclear priorities

• A politics player (we don't have room for that)

What Success Looks Like (Year 1 & Beyond)

In the First 90 Days

• You've mapped the technical landscape: where the bottlenecks are, which systems are fragile, where we're overspending on compute or tokens

• You've identified 3-5 technical leaders on the team and are working with them directly

• You've shipped at least one meaningful piece of code (not ceremonial—something that improves the system)

• You've had multiple technical sparring sessions with Manick where you've pushed back on decisions and earned trust through depth

In Year 1

• Engineering velocity has noticeably increased: faster cycle times, more confident deployments, fewer surprise production incidents

• The quality bar has risen: evals are more rigorous, code reviews are sharper, but you've done this through mentorship, not rules

• You've solved at least one major technical problem that was blocking the organization: a scaling bottleneck, a reliability issue, a cost problem

• You've hired 5-10 exceptional engineers who all say "I came to work for [you], not the company"

• You've reduced technical debt in critical systems while shipping features

• The organization runs tighter, faster, with less friction

In Years 2-3

• We've scaled to $75-100M ARR with engineering quality improving, not degrading

• You've developed 3-4 technical leaders who could step into your role

• The agentic systems we've pioneered are industry-leading—harder to copy than our product features

• We've solved the hard technical problems that are currently blocking us, and we're now solving the next level of hard problems

• You own a significant piece of equity in a profitable, high-growth company