Digital AI Staff Engineer

Join to apply for the Digital AI Staff Engineer (Contract-to-hire) role at LextechAbout the RoleDigital AI Staff EngineerYou bring digital staff to life using the latest AI tools and techniques. You'll design intelligent agents and create the tools they use to access systems and data in the broader world. You know how to combine AI logic, real-world integrations, and reliability practices so digital staff can perform like seasoned teammates.About YouRequired Technical SkillsAI FundamentalsUnderstands how LLMs interpret prompts and contextBasic understanding of agent memory and context windowsAwareness of the strengths and limits of different modelsBuilding Core AgentsComfortable modeling business logic as workflows via promptsDesigns for non-technical end usersDesigns fallbacks, retries and escalationsDesigns incorporate logging and monitoringBuilding ToolsComfortable building integrations with REST API endpointsFamiliar with authentication and authorization flows (OAuth 2, API Keys, etc)Comfortable scripting in Python and/or JavaScript (parsing, transforming and mapping data)CollaborationConsults documentation before diving inDocuments new agent & tool patternsAbout UsAt Lextech, we created AIWhy to help companies scale with digital staffers — AI-powered teammates that take on real work. Our digital staff plug into existing teams, systems and processes, delivering business outcomes quickly without the heavy lift of traditional tech projects. We're obsessed with making AI practical, reliable, and results-driven so our clients can focus on what matters: growing their business.About the JobResponsibilitiesBuild digital staff that deliver results: design, prompt, and configure AI agents that perform real work and drive business outcomes in n8n, RelevanceAI and other platforms.Translate needs into solutions: collaborate with Outcome Architects and teammates to understand customer goals and translate them into working agents.Keep digital staff running smoothly: test, tune, and iterate quickly to maintain performance and reliability. Tune digital staff prompts and rework staff to address any issues and tuning output results.Extend agent capabilities: build and maintain integrations and tools that give agents access to data and systems they need.Performance PrinciplesLearning sponge: the environment is changing faster than anything before with lots to learn. Exciting stream of learning with tinkering, research and implementation.Prompt whisperer: clearly craft AI agent prompts that utilize tools to get the job done and handle exception cases and feedback loops cleanly. You test, measure, and deliver.Know the why: understand what customer success looks like for each staff member and ensure that's the focus, not the tech.No surprises: you communicate early and clearly, keeping your team and leaders informed before problems grow.OutcomesDigital staff are humming: customized staff are running smoothly with minimal intervention and delivering on its defined goals.Quality matters: digital staff intentionally do what they're supposed to, not occasionally or by accident. All builds are easy to understand, maintain, and reuse.Fast turnarounds: new agents and updates move from concept to working versions quickly.Continuous improvement: agents are tuned and refined based on feedback, performance data, and client outcomes.First 6 MonthsFirst agents live and delivering measurable results: you've built and tuned digital staffers that clearly demonstrate business value.Agent library growing: several reusable agent or tool patterns documented and shared with the team.Speed and reliability proven: quick turnaround and consistent delivery on roadmap items.New integrations working in production: agents successfully using external data and systems through tools you've built.Deal-BreakersDefaulting to no: instead of exploring options and paths to yes (unless no is the right answer after vetting).Keeping others in the dark: if leadership or the team has to chase you for status, it is already broken.Fire and forget doesn't work: staff functioning needs to be validated and regularly observed including adding key checkpoints into ops logging. Something working once is not a sign of being done.Focusing on tech instead of results: clients care about outcomes, not tools. #J-18808-Ljbffr