AI Engineer

Scope Applied AI; San Francisco (CA); Hybrid; USD 180,000 - 210,000; Full time

Job summary

Scope Applied AI in San Francisco seeks an AI Engineer to build reliable AI software for real business operations. You will help move prototypes to deployed systems and own implementation details while shaping where AI adds value versus traditional software.

The role blends hands-on software engineering with applied AI, including tool integrations, data connections, and monitoring for real-world performance. Hybrid work in San Francisco is offered.

Qualifications

• Track record of shipping AI-powered software used by real users and ownership of significant technical decisions.
• Strong backend engineering in Python and/or TypeScript, including API design, data modeling, asynchronous processing, and automated testing.
• Hands-on experience designing multi-step AI or agent workflows with tool use, explicit state management, retries, idempotency, and recovery from partial failures.
• Experience building evaluations from representative tasks and data, analyzing failure patterns, and using regression checks to make release decisions.
• Ability to diagnose failures across model behavior, prompts, retrieval, application code, and external systems, and turn that diagnosis into a durable fix.
• Evidence of improving reliability, latency, or cost in a deployed system, supported by measurements and concrete technical tradeoffs.
• Ability to discuss systems you personally built in depth: architecture choices, failure modes, testing approach, and what changed after deployment.
• Valid authorization to work in the United States.

Responsibilities

• Build and deploy LLM-powered workflows and agents, including tool integrations, state management, and failure handling.
• Connect AI systems to enterprise APIs, databases, documents, and existing operational tools.
• Build retrieval and knowledge workflows when the use case requires them.
• Create evaluation datasets and checks that measure task completion, output quality, and regressions.
• Improve reliability, latency, and cost through testing, instrumentation, and production monitoring.
• Work with AI Strategists and engineers to turn workflow requirements into clear technical decisions and usable systems.
• Document technical decisions and contribute reusable engineering components and practices.

Job description

We're looking for an AI Engineer to build reliable AI software for real business operations. You will work with the founding team and the engineering team to take an operating problem from prototype to a deployed system that people can depend on.

The role combines hands‑on software engineering with applied AI: integrating models and tools, designing agent workflows, connecting enterprise data and APIs, and measuring how the system behaves under real conditions. You will own implementation details while helping decide where AI adds value and where conventional software is the better choice.

This role has a hybrid work arrangement in San Francisco.

• Build and deploy LLM‑powered workflows and agents, including tool integrations, state management, and failure handling.
• Connect AI systems to enterprise APIs, databases, documents, and existing operational tools.
• Build retrieval and knowledge workflows when the use case requires them.
• Create evaluation datasets and checks that measure task completion, output quality, and regressions.
• Improve reliability, latency, and cost through testing, instrumentation, and production monitoring.
• Work with AI Strategists and engineers to turn workflow requirements into clear technical decisions and usable systems.
• Document technical decisions and contribute reusable engineering components and practices.

• A track record of shipping AI‑powered software used by real users, with personal ownership of significant technical decisions and production outcomes.
• Strong backend engineering in Python and/or TypeScript, including API design, data modeling, asynchronous processing, and automated testing.
• Hands‑on experience designing multi‑step AI or agent workflows with tool use, explicit state management, retries, idempotency, and recovery from partial failures.
• Experience building evaluations from representative tasks and data, analyzing failure patterns, and using regression checks to make release decisions.
• Ability to diagnose failures across model behavior, prompts, retrieval, application code, and external systems, and turn that diagnosis into a durable fix.
• Evidence of improving reliability, latency, or cost in a deployed system, supported by measurements and concrete technical tradeoffs.
• Ability to discuss systems you personally built in depth: architecture choices, failure modes, testing approach, and what changed after deployment.
• Valid authorization to work in the United States.

• Familiarity with retrieval systems, vector databases, and evaluation tooling.
• Experience with enterprise integrations, asynchronous workflows, or human‑in‑the‑loop systems.
• Experience with cloud infrastructure, CI/CD, and observability.
• Experience working directly with users or clients to improve a deployed system.