Agentic AI Engineer

Provn, Inc.; Seattle (WA); Hybrid; USD 150,000 - 190,000; Full time

Job summary

Envorso Sports is hiring an Agentic AI Engineer to scale our fan-facing platform. You will build agents turning fan and ticketing data into revenue, and ship front-end surfaces across Next.js web and mobile apps so agents reach fans and marketers.

Reporting to product with daily collaboration from the tech lead and Argentina-based engineering team, you’ll define guardrails, tooling, and end-to-end lifecycle management in a fast-paced startup environment.

Responsibilities

• Design and ship agentic product features: fan scoring, segmentation, and an offers engine that proposes campaigns instead of just reporting on fan behavior
• Build the front-end for these features on both web (Next.js) and mobile, so agent output is visible to marketers or fans
• Build and own our internal agent tooling: guardrails, PR-review agents, spec-driven test generation, and contracts synchronization tooling
• Set the standard for model selection, cost, and evaluation across the team, and mentor engineers adopting agentic workflows
• Own agent quality: catch shallow or overfit AI output and make failures visible, not silent
• Own the full lifecycle of what you build — design, evaluation, deployment, monitoring, iteration — without an ML platform team

Job description

About Envorso Sports

Envorso Sports is building the platform that runs a professional sports team's entire fan-facing business — ticketing, website, mobile app, and eventually commerce, CRM, and marketing — from one system instead of six disconnected vendors.

Today, most teams run a fragmented stack: a website vendor, a ticketing provider, a separate mobile app, CRM tools, marketing automation, and game-day engagement solutions bolted together. That creates a disconnected fan journey and leaves revenue on the table.

Our platform is live today with the Seattle Seawolves of Major League Rugby, with expansion planned across the league and into additional professional sports organizations. Ticketing, website, and mobile app are shipped and in production with a paying customer. CRM, marketing automation, and AI-driven personalization are near-term roadmap, not yet built — we want candidates to know exactly where the product stands before they join.

The Role

We're hiring an Agentic AI Engineer to scale the Envorso sports platform, both as a product capability and as how our engineering team builds.

On the product side, you'll build agents that turn fan and ticketing data into revenue: an agent that flags a stale cart and proposes a win-back offer, or builds a segment a marketer can act on without an analyst in the loop — rather than a dashboard that just reports what happened.

On the engineering side, you'll build our internal agent tooling — guardrails on what coding agents can touch, PR-review agents that check against real specs, and automated checks that keep frontend and backend in sync as API contracts change — and set the standard the rest of the team works against.

Given our stage, this isn't a backend/ML-only role: you'll also ship the front-end surfaces that expose this work, across both our web app (Next.js) and our mobile app, so the agent features you build actually reach fans and marketers rather than living behind an API.

You’ll report into product and work daily with our tech lead and engineering team, who are mostly Argentina-based. Expect real overlap-hours collaboration.

What You’ll Do

• Design and ship agentic product features: fan scoring, segmentation, and an offers engine that proposes campaigns instead of just reporting on fan behavior
• Build the front-end for these features on both web (Next.js) and mobile, so agent output is something a marketer or fan sees and acts on, not just an API response
• Build and own our internal agent tooling: guardrails, PR-review agents, spec-driven test generation, and agents that keep frontend/backend in sync as contracts change
• Set the standard for model selection, cost, and evaluation across the team, and mentor engineers adopting agentic workflows
• Own agent quality: catch shallow or overfit AI-generated output, define what "good" looks like, and make failures visible instead of silent
• Own the full lifecycle of what you build — design, evaluation, deployment, monitoring, iteration — without an ML platform team to hand off to

What This Role Isn’t

Not a research role — we need things that work in production against real fan and ticketing data. Not a role with a fully scoped backlog handed to you each sprint.

Work Authorization: Must be authorized to work in the US. No employer-sponsored visas for this role.