Senior Software Engineer

Evolve Group; New York (NY); On-site; USD 200,000 - 400,000; Full time

Job summary

Evolve Group in the United States is seeking a Senior Software Engineer / Research Engineer focused on AI systems. The role sits between distributed systems, backend engineering, and applied AI to build the execution layer behind production AI agents.

Location options include San Francisco or New York, with base compensation of $200K to $400K plus competitive equity. The team is expanding, and early hires will help shape architectural direction and production-ready infrastructure.

Qualifications

• Experience owning technically complex systems from design through production.
• Strong ability to debug difficult production problems.
• Experience improving reliability, performance, and observability.

Responsibilities

• Design and build execution frameworks for production AI agents.
• Build systems for routing and orchestrating multiple models.
• Develop control-plane logic for planning, execution, and tool invocation.
• Build reliable distributed systems handling large volumes of real-time interactions.
• Optimize systems for latency, reliability, and production correctness.
• Diagnose real-world agent failures and identify root causes.
• Build feedback loops that improve agent behaviour over time.
• Develop online experimentation and A/B testing infrastructure.
• Contribute to offline evaluation and simulation systems.
• Improve observability, testing, and production monitoring.
• Work on real-time systems with strict latency requirements.
• Continuously rethink architecture as model capabilities improve.

Skills

Distributed systems

Backend engineering

AI infrastructure

Debug production problems

Job description

Senior Software Engineer / Research Engineer, AI Systems

Location: San Francisco or New York

Compensation: $200K to $400K base + competitive equity

About the Role

A rapidly scaling AI company is building out a new engineering team focused on the infrastructure that governs how production AI agents reason, execute, use tools, and interact with users.

This is a highly technical role sitting between distributed systems, backend engineering, and applied AI.

You will work on the execution layer behind production agents, including model routing, planning, tool orchestration, control systems, experimentation, and real-time infrastructure.

The problems are highly experimental and change quickly as underlying model capabilities evolve.

What You'll Work On

• Design and build execution frameworks for production AI agents
• Build systems for routing and orchestrating multiple models
• Develop control-plane logic for planning, execution, and tool invocation
• Build reliable distributed systems handling large volumes of real-time interactions
• Optimize systems for latency, reliability, and production correctness
• Diagnose real-world agent failures and identify root causes
• Build feedback loops that improve agent behaviour over time
• Develop online experimentation and A/B testing infrastructure
• Contribute to offline evaluation and simulation systems
• Improve observability, testing, and production monitoring
• Work on real-time systems with strict latency requirements
• Continuously rethink architecture as model capabilities improve

What They're Looking For

• Experience owning technically complex systems from design through production
• Strong ability to debug difficult production problems
• Experience improving reliability, performance, and observability
• Comfortable operating in ambiguous environments where the solution is not already known
• Strong technical judgement and ability to iterate quickly
• Experience working closely with research, infrastructure, and product teams
• Evidence of high technical trajectory and increasing ownership

Particularly Relevant Experience

Strong candidates may have worked on:

• Distributed systems
• Backend platforms
• AI infrastructure
• Agent systems or execution frameworks
• Model serving or inference infrastructure
• Search or retrieval systems
• Real-time systems
• Control planes
• Evaluation infrastructure
• High-performance infrastructure
• Quantitative or technically demanding engineering environments

Direct agent experience is valuable, but not mandatory.

Exceptional engineers from strong systems, infrastructure, startup, or quantitative backgrounds can also be highly relevant.

The Engineering Problems

The core challenge is not simply calling an LLM.

It is building everything around the model that determines whether an intelligent system actually works reliably in production.

That includes:

Execution

Determining which workflows, models, and tools should run and in what order.

Orchestration

Coordinating multiple models and components across complex tasks.

Reliability

Ensuring agent behaviour remains predictable and correct across millions of interactions.

Making intelligent systems respond quickly enough for real-time applications.

Understanding why systems fail, testing improvements, and measuring whether changes actually work.

Continuously redesigning infrastructure as frontier models improve.

Why Consider It

Frontier technical problems

You will work on problems where established engineering patterns do not always exist yet.

The team is being built out now, creating substantial scope for individual engineers to influence architecture and technical direction.

High technical bar

The team is targeting engineers from strong startups, infrastructure organizations, quantitative environments, and leading technical institutions.

Production impact

The systems you build directly determine how AI agents reason, take actions, and behave in live environments.

Rapid growth

The team plans to add approximately 6 to 8 engineers, creating opportunities for early hires to take on significant scope as it scales.

Compensation

Base salary:

$200K to $400K

Additional compensation:

Competitive equity

Typical leveling ranges approximately from:

• Mid-level: $250K to $280K
• Senior: $300K to $330K
• Staff: $350K to $400K

Final compensation depends on experience, technical depth, and level.