Staff Software Engineer, Applied AI
PrizePicks is building an Applied AI engineering team focused on turning modern ML/LLMs into reliable, scalable product capabilities. This is a senior individual contributor role for someone who has operated on an Applied AI/Product AI team before - someone who can build, ship, and iterate quickly, while also raising the bar for engineering rigor and judgment in an agent-driven SDLC.
In the near term, you'll help build the agent platform and internal tools, and partner with application engineering teams to operationalise AI across the software development lifecycle. Over time, the Applied AI team will increasingly shape product roadmap and strategy for AI-powered experiences - and increasingly, how the rest of engineering works day to day.
- Build our agent platform and tooling. Design and implement primitives like orchestration, tool/function calling, evaluation harnesses, prompt/version management, tracing/observability, and safety/guardrails.
- Support patterns like retrieval-augmented generation (RAG), structured extraction, and production inference workflows.
- Operationalise AI in the SDLC. Work with application engineering teams to embed AI into day-to-day engineering workflows (code review assist, test generation, incident support, developer copilots), with clear quality gates and measurable impact.
- Own the judgment call of what to delegate to agents vs. keep as human work - scope tasks so agents are set up to succeed, and adapt that boundary as team skill levels and tooling maturity vary.
- Reason from the trace, not just the output. Given an agent run, diagnose whether it did its job well from the trace itself - not just the final diff. Debug agent behavior the way you'd debug a distributed system: inputs, intermediate steps, failure points. Treat AI output as a claim to verify, not a result to trust - including tests the agent writes for its own code.
- Ship applied AI features end-to-end. Own projects from prototype → production: data needs, system design, model/vendor selection, rollout plans, monitoring, and iteration. Partner closely with product, design, and data/ML stakeholders to deliver customer-facing outcomes (not just demos).
- Diagnose and evangelise AI adoption org-wide. Assess where AI adoption is working and where it isn't - identify bottlenecks and failure modes across teams, not just within your own code. Build what's missing, teach what's already built, and make the case for adoption to skeptical or hesitant engineers.
- Mentor and set engineering standards. Establish best practices for reliability, evaluation, incident response, privacy/security, and "how we build AI here." Coach other engineers through design reviews, pairing, and pragmatic technical leadership - including how to critically review AI-generated code rather than rubber-stamp it.
- 6+ years of professional software engineering experience (or equivalent), including shipping production systems (required).
- Direct experience on an Applied AI / product AI team building LLM- or ML-powered features (agents/tooling strongly preferred) (required).
- Demonstrated judgment in agent-driven development: knowing what to delegate, how to scope/context an agent, and how to verify its output before trusting it - this is as much a hiring bar here as raw coding ability (required).
- Strong backend/system design skills: APIs, distributed systems, queues/workflows, observability, and performance (required).
- Comfort navigating ambiguity and driving outcomes with cross-functional partners (product, design, data/ML, platform) (required).
- Track record of mentoring peers and leveling up engineering quality - including helping others build critical judgment around AI-assisted output, not just AI literacy (required).
- Experience building "platform" capabilities for other engineers (SDKs, internal frameworks, developer tooling) (nice-to-have).
- Practical experience with embeddings, search/retrieval, evaluation methodologies, and model monitoring (offline + online) (nice-to-have).
- Experience building AI-powered tooling/products like agents and assistants at scale (nice-to-have).
- Familiarity with inference constraints (latency, cost, caching), vendor/model tradeoffs, and deployment patterns (nice-to-have).
- Company-subsidised medical, dental, & vision plans.
- 401(k) plan with company match.
- Annual bonus.
- Flexible PTO to encourage a healthy work/life balance (2 weeks STRONGLY encouraged!).
- Generous paid leave programs, including 16-week paid parental leave and disability benefits.
- Workplace flexibility and modern work schedules focused on getting the job done, not hours clocked.
- Company-wide in-person events and team outings.
- Lifestyle enhancement program.
- Company equipment provided (Windows & Mac options).
- Annual performance reviews with opportunities for growth and career development.
PrizePicks is a daily fantasy sports operator headquartered in Atlanta, Georgia. Its app lets players predict whether athletes will beat or fall short of projected statistics, and it has added prediction-market contracts regulated by the US Commodity Futures Trading Commission. In January 2026 the lottery and gaming group Allwyn completed the purchase of a 62.3% majority stake, with PrizePicks continuing as a standalone brand.