Software Engineer, Agent Systems

Serval; San Francisco (CA); On-site; USD 200,000 - 325,000; Full time

Job summary

Serval is seeking a Software Engineer, Agent Systems to build the core systems powering AI agents in our platform. You’ll work on orchestration, runtime, grounding, and evaluation infrastructure to enable agents to reason, take action, and operate reliably in enterprise environments.

You will help shape how Serval agents behave across help desk, automation, and knowledge workflows, with systems that scale to thousands of users at large enterprises.

Qualifications

• 4+ years of hands-on experience building software products in production.
• Experience designing AI-powered systems and agent frameworks.
• Strong fundamentals in software engineering and reliability.
• Ability to work cross-functionally in a fast-moving startup.

Responsibilities

• Build end-to-end agent based product functionality.
• Build the foundational systems behind Serval’s agent platform, including orchestration, runtime, and core primitives for how agents reason, act, and interact with users and systems.
• Design and evolve the agentic loop, helping define the abstractions needed to make agents steerable, trustworthy, adaptive, and production-ready.
• Improve retrieval and grounding systems so agents can provide accurate, well-supported answers and actions using the right organizational context.
• Build evaluation frameworks and feedback loops that help us measure, debug, and continuously improve agent quality over time.
• Partner closely with product to bring new agent capabilities into customer-facing workflows and product experiences.
• Help establish technical standards and best practices for building reliable, high-performance agent systems at Serval.

Skills

Software engineering

AI-powered systems

Agent frameworks

LLM-based applications

Cross-functional collaboration

Education

CS degree or equivalent experience

Tools

Go

TypeScript

Job description

Who We Are

Serval is an AI-native automation platform transforming how enterprises operate. We build intelligent agents that understand real-world workflows and execute them end-to-end — replacing manual processes and rigid legacy systems with adaptive, learning software. Founded in early 2024, Serval is already trusted by companies like Fox, Notion, Perplexity, Vercel, and Brex to automate high-volume, high-friction operational work across their organizations. At the core of Serval is an agentic AI platform that turns natural language into production-grade workflows. Our agents don’t just respond to requests — they reason, take action across systems, and continuously improve with usage. What began with operational use cases has quickly evolved into a horizontal AI automation layer used across IT, HR, Finance, Security, Legal, and Engineering. Our mission is to eliminate repetitive, manual work across the enterprise and give teams leverage through intelligent automation. Long term, we’re building the universal AI operations layer — a system of agents that sits across business functions and runs the workflows that keep modern companies moving. We’re backed by leading investors including Sequoia Capital, Redpoint Ventures, Meritech, First Round, General Catalyst, Elad Gil, and others.

Role Overview

As a Software Engineer, Agent Systems at Serval, you’ll build the core systems that power the AI agents at the center of our platform. You’ll work on the orchestration engine, runtime, retrieval and grounding systems, and evaluation infrastructure that enable agents to reason, take action, and operate reliably in real-world enterprise environments. Your work will help define how Serval agents behave across help desk, automation, and knowledge workflows — and the systems you build will power experiences used by a significant share of employees at some of the world’s largest companies.

What You’ll Do

• Build end-to-end agent based product functionality.
• Build the foundational systems behind Serval’s agent platform, including orchestration, runtime, and core primitives for how agents reason, act, and interact with users and systems.
• Design and evolve the agentic loop, helping define the abstractions needed to make agents steerable, trustworthy, adaptive, and production-ready.
• Improve retrieval and grounding systems so agents can provide accurate, well-supported answers and actions using the right organizational context.
• Build evaluation frameworks and feedback loops that help us measure, debug, and continuously improve agent quality over time.
• Partner closely with product to bring new agent capabilities into customer-facing workflows and product experiences.
• Help establish technical standards and best practices for building reliable, high-performance agent systems at Serval.

What You’ll Need

• Strong software engineering fundamentals and experience building production-grade systems.
• Experience building AI-powered systems, agent frameworks, or LLM-based applications in production.
• 4+ years of hands-on experience building software products and systems in production environments.
• Experience designing systems from first principles, with an eye toward reliability, scalability, and maintainability.
• Excitement about building at the frontier of AI products, with a strong point of view on what makes agents useful in the real world.
• High agency and comfort operating in a fast-moving, high-autonomy startup environment.
• Strong communication skills and the ability to work cross‑functionally.
• Degree in Computer Science or a related technical field, or equivalent professional experience.

Nice to Have

• Experience with retrieval systems, grounding strategies, evaluation infrastructure, and other components of modern agent architectures.
• Familiarity with technologies in our stack, including Go and TypeScript.
• Early-stage startup experience or track record of zero-to-one product development.

What we offer

• Impact: Be a key player in shaping the success of our product and company.
• Growth: Build a fundamentally new AI product offering with the support of our experienced team and investors. Grow rapidly with the company.
• Culture: Join a culture that values innovation, ownership, accountability and fun.

Compensation Range: $200K - $325K