Product Owner

Credibly•Southfield, MI

2d•Hybrid

About The Position

Credibly runs AI models, LLM-powered deal and document review, rules-based decisioning, and a broad set of third-party data and partner integrations in production. Keeping them accurate, reliable, and improving takes someone who lives in the details. As Product Owner, you will own the feedback loop between the people who use our platform and the team that builds it. You will gather what our merchants, partners, and internal teams experience, turn it into clear, prioritized work for an agile engineering team, and follow every change through to release. Then you will measure the result, whether that is time saved, fewer errors, or faster funding, so each improvement is backed by data. You will partner closely with Product Management, which owns the roadmap and overall outcomes.

Requirements

2+ years as a Product Owner, Business Analyst, or technical product role on an agile software team
Experience using AI tools in your day-to-day work, such as drafting user stories, summarizing feedback, or analyzing data, and good judgment about when to double-check the results
Excellent story-writing skills; your acceptance criteria are specific enough to test against
Proficiency in SQL for investigating issues and validating data
Hands-on experience testing APIs (e.g., Postman) and working with JSON payloads, webhooks, and field-level data mapping
Practical experience with LLM tools, including iterating on prompts and testing outputs in a structured, repeatable way
Experience writing and executing test cases and running UAT
Strong command of Jira/Confluence or similar tools
Excellent story-writing skills; your acceptance criteria are specific enough to test against
Meticulous attention to detail and clear written communication
Bachelor's degree or equivalent practical experience

Nice To Haves

Experience in small business lending, MCA, credit underwriting, KYB/KYC, or fraud operations
Experience with third-party integrations common in lending, such as bank-data providers like Plaid, credit reporting agencies, and KYC/KYB verification vendors
Experience with BI tools such as Metabase
Experience working with data science teams on products that use machine learning models, such as writing requirements, testing model changes, or tracking model performance after release
Using genAI to perform basic Python or scripting ability for data checks and test automation

Responsibilities

Own the product feedback loop.
Connect Operations, Engineering, and Data Science: bring what underwriters and sales see in production (overrides, exceptions, defects) to the teams who can fix it, make sure model and rules changes are understood and adopted by Operations, and report back on the impact.
Run the agile rhythm. Lead backlog refinement, sprint planning, and sprint reviews, and keep work visible and moving in Jira.
Translate policy into logic. Work with Underwriting to turn procedures and policy into decision tables, rule logic, and data requirements engineering can implement without guesswork.
Spec integrations end to end. Define API contracts, field mappings, error handling, retries, and fallback behavior for data vendors and funding partners; test in vendor sandboxes; and own data mapping for partner data tapes and reporting.
Test model and prompt changes. Maintain evaluation datasets, run evaluations against the current baseline, document the results, and sign off (or send back) before anything reaches production.
Own acceptance and UAT. Write and execute test cases, coordinate user acceptance testing with business users, and maintain regression coverage.
Watch production closely. Monitor dashboards and alerts for model accuracy, integration error rates, vendor outages, and latency; triage defects; investigate with SQL; and write bug tickets that engineers can act on immediately.
Keep documentation current. Maintain the rules catalog, model and prompt change logs, integration docs, release notes, FAQs, and training materials, and lead training sessions with business users.
Own the backlog. Write and prioritize epics and user stories with precise acceptance criteria for model, prompt, rules, and API work, and serve as engineering's go-to decision maker during the sprint.