AI Outcomes Associate
hace 24 días
Dallas
Job DescriptionAbout UpSmith UpSmith builds agentic AI systems that help skilled-trades businesses win—by converting more demand into revenue without adding payroll. Our flagship platform, Jenny, deploys autonomous AI agents that engage customers, surface revenue opportunities, and execute end-to-end workflows directly inside systems like ServiceTitan. We are entering a phase where AI capability alone is no longer the bottleneck. Rather, the bottleneck is product outcomes: deploying agents into messy, real-world environments and making them work—reliably, measurably, and at scale. This role exists to own that problem, and associated outcomes with it. The Role The AI Outcomes Associate sits at the intersection of: • Applied AI & agent design, • Product judgment & abstraction discipline This is not a traditional CSM role. It is a front-line, ownership-heavy role for someone who wants to see their work running in production—creating real economic value. What You’ll OwnAI Outcomes & Customer Impact • Own the end-to-end success of AI agent deployments for our customers, • Define, track, and improve leading and lagging indicators of agent success (conversion, bookings, revenue lift, customer trust), • Diagnose failures across data, prompts, tooling, coordination logic, and customer workflows, • Act as the single accountable owner for “Is this working—and why?”, • Work hands-on with AI agents in production:, • Prompting and behavior design, • Guardrails, fallbacks, and human-in-the-loop escalation, • Partner with core engineering on deeper system improvements surfaced from the fieldProduct & Feedback Loop, • Translate customer reality into clear product insights and roadmap input, • Identify when problems are:, • Agent behavior issues, • Product abstraction issues, • Customer workflow mismatches, • Help decide what not to build by grounding decisions in outcomes, not hypotheticals What Success Looks Like (6–12 Months) • Customers expand usage because the configuration, deployment, and ongoing maintenance/support of AI agents measurably drive revenue at clients, • Agent deployments become faster, more repeatable, and less bespoke, • Product and engineering velocity increases due to high-signal field feedback, • You are trusted internally to own some customer outcomes with minimal oversightOur Values (Non-Negotiable), • Mission first. Deliver magic for the builders we serve., • Speed wins. Execute efficiently with bias to action., • Be an owner. Raise the bar and take pride in the work. Why This Role Matters UpSmith is moving from “Can we build agents?” to “Can we deploy agents that win?” This role is where that question gets answered—every day, with real customers, real revenue, and real accountability. Requirements What You Must HaveCore Capabilities • Strong analytical and systems thinking skills, • Comfort working across codebases, tooling, product, and customer conversations, • Ability to operate in ambiguity and make sound judgment calls quickly, • Bias toward action with a high bar for qualityTechnical Fluency, • Working knowledge of:, • LLMs and their failure modes, • Prompt engineering, • SQL and data visualization tools (i.e. Tableau, Retool, Metabase), • You don’t need to build solutions from scratch, but you should have the capability and reasoning to scale and help plan for the future on our current systemsCustomer & Ownership Mindset, • You take responsibility when things don’t work, • You care deeply about real-world impact, not demos, • You are comfortable telling hard truths—to customers and teammatesLogistics, • Willingness to travel frequently to customer sites, • Comfortable operating in high-growth, high-expectation environments Bonus (Not Required, but Powerful) • Management consulting, product ops, or startup experience, • Strong SQL & data visualization skills, • Prior experience deploying AI or automations into production environments, • Familiarity with skilled trades, home services, or operationally complex businesses, • Working knowledge of APIs, data pipelines, and production constraints