SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Unify is building an AI-native outbound platform where agents and sellers collaborate to help revenue teams identify buyers, enrich account data from 40+ sources, and execute personalized multi-channel outreach from a single workspace. The platform delivers up to 90% time savings on list building and 19% higher output per rep.
As an Applied AI Engineer, you will own the end-to-end development of new agent capabilities from prototype to production. You'll design and build agent harnesses—the execution loops, tool interfaces, context management, memory systems, and error recovery mechanisms that enable agents to complete complex workflows reliably.
Key responsibilities include:
- Building agent harnesses that design execution loops, tool interfaces, context, and memory for reliable complex workflow completion
- Shipping AI products end-to-end by translating ambiguous customer problems into working features, connecting agents to Unify's data, integrations, and workflow infrastructure
- Making quality measurable by building evaluations from real customer tasks and using traces and failure analysis to improve agent behavior
- Building for production by owning reliability, latency, and cost; making long-running tasks resumable; handling tool failures; and providing users visibility and control
- Experimenting with new capabilities by testing models, prompting strategies, retrieval approaches, and orchestration patterns, then measuring and shipping improvements
- Shaping the AI foundation by building reusable tools and engineering practices that help the team ship faster, in partnership with product and design
You'll work closely with product, design, and engineering teams, moving quickly from prototype to customer usage while using evaluations and real execution traces to continuously improve the product.
REQUIREMENTS:
- 3+ years of software engineering experience with hands-on experience shipping LLM applications or agents to production
- Built systems with tool calling, multi-step execution, and evaluations; understand how software around a model affects its capabilities
- Strong programmer in TypeScript or Python with backend systems and API integration experience; comfortable owning features across the stack
- Empirical thinker who defines success, runs experiments, inspects failures, and uses evidence to decide what to build
- Deep customer focus; ability to turn high-level product ideas into deployed solutions while balancing speed with sound engineering judgment
- Self-starter who learns quickly, collaborates generously, and thrives with ambiguity
