Cognitive Systems Engineer

Snap Finance • West Valley City, UT

2d

About The Position

At Snap Finance, we believe everyone deserves access to the things they need, regardless of credit history. Since 2012, we've used data, machine learning, and a more human approach to create flexible financing solutions that help people move forward. We're proud of our inclusive, supportive culture, built on empowering our customers, partners, and team members alike. When our people thrive, so does our innovation. If you're looking to make an impact and grow with a team that values you, come join us! This role involves designing and building agentic analytics systems that can reason over data, generate insights, and take action with minimal human intervention. The engineer will define and manage the context lifecycle for agents, write high-quality agent specifications, optimize token efficiency, latency, and cost across multi-agent workflows, and develop automated orchestration frameworks. Key responsibilities also include implementing evaluation and quality control loops, translating business questions into agent-executable analytics workflows, building and maintaining data pipelines, and partnering with analytics, engineering, and product teams to embed these systems into business processes. Continuous experimentation with new models, tooling, and architectures is also expected.

Requirements

• Strong foundation in data analysis and/or data science (SQL, Python, statistical thinking, experimentation)
• Experience building production-grade data systems or ML systems (pipelines, APIs, orchestration frameworks)
• Working knowledge of LLMs and agentic architectures (prompt design, tool use, memory strategies, evaluation techniques)
• Ability to think in systems, not just models—understanding how components interact, fail, and scale
• Experience or strong interest in automation and autonomous workflows
• Comfort with ambiguity and a bias toward building, testing, and iterating quickly
• Clear, structured thinking with the ability to write precise specifications for both humans and machines
• Strong engineering discipline around versioning, testing, observability, and performance optimization
• Curiosity about the future of analytics and a drive to move beyond dashboards into decision systems

Responsibilities

• Design and build agentic analytics systems that can reason over data, generate insights, and take action with minimal human intervention
• Define and manage the context lifecycle for agents (ingestion, memory, retrieval, pruning, and evaluation)
• Write high-quality agent specifications (prompts, tools, workflows) that are modular, testable, and production-ready
• Optimize token efficiency, latency, and cost across multi-agent workflows without sacrificing output quality
• Develop automated orchestration frameworks for chaining agents, handling fallbacks, and managing state across complex tasks
• Implement evaluation and quality control loops (automated scoring, human-in-the-loop, and self-healing systems)
• Translate ambiguous business questions into structured, agent-executable analytics workflows
• Build and maintain data pipelines and feature layers that support real-time and batch agent reasoning
• Partner with analytics, engineering, and product teams to embed agentic systems into real business processes
• Continuously experiment with new models, tooling, and architectures to push system capability forward

Benefits

• Generous paid time off
• Competitive medical, dental & vision coverage
• 401K with company match for US
• Company-paid life insurance
• Company-paid short-term and long-term disability
• Access to mental health and wellness resources
• Company-paid volunteer time to do good in your community
• Legal coverage and other supplemental options
• A value-based culture where growth opportunities are endless

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed
• Number of Employees: 251-500 employees