Product Manager

Who we are

At Fido, we are building the future of finance in Africa powered by advanced technology, data-driven decision making, and bold thinking.

Through AI, cutting-edge data science, and automation, we're redefining how people access and experience financial services. Our goal is to make finance simple, smart, and accessible giving everyone the confidence to take charge of their financial story.

Joining Fido is an opportunity to drive real impact, solve meaningful problems, and contribute to building a future where millions have the tools to create, grow, and thrive.

Job summary

We are a lender, not an app sitting on someone else's balance sheet. We underwrite, we disburse, we collect, and we take the loss when we're wrong. So every product decision here is also a credit decision, an onboarding flow that converts better but lets in worse borrowers is not an improvement.

We also run on agents. Not as a pilot. Product, design, and engineering work through a system we built ourselves. It writes specs, runs analyses, drafts designs, and moves tickets.

You own a domain of the business, customers, numbers, and a line in the P&L, with your name on it. You decide what winning looks like, argue for the resources, and answer for the outcome.

Product owners sit inside the engineering team. There is no product org, no roadmap document, and no analyst who owns your numbers for you.

What that means day to day:

• Your numbers. Conversion, repayment, cost to acquire, cost to collect, contribution per customer. You defend them in front of leadership. You will argue about pricing, limits, and risk cut-offs and sometimes give up conversion to buy credit quality.
• The whole thing. Problem, data, decision, build, launch, measurement, next decision. There is no point where it stops being yours.
• Your agents. You set the intent, give them the domain knowledge that makes them correct, and throw out what comes back wrong. You'll also write shared skills that become how the whole company does a piece of work. Your leverage is judgment and context, not output.
• The argument. Risk, collections, marketing, compliance, and engineering all have a say in your domain. Getting them to yes is your job.
• The constraints. Low-literacy users. Cheap Android phones. Expensive data. Bad networks. Four regulators. Anything designed for a laptop fails here.

What we expect:

Month 1: You pull your own data and know your baseline.

Month 3: You shipped something that moved a number, and you can show the before and after.

Month 6: When leadership wants to know what's happening in your domain, they ask you, not the analyst.

Who we're looking for

We care what you have owned, not how long you've held a title.

• You have owned a number. Somewhere there's a metric that was yours, that moved, and you can explain why. A PM title is optional, some of our best owners came from analytics and engineering.
• You are numerate on your own. You write SQL, or you learn it in the first week. You build the view you need instead of waiting for a dashboard. You do not present numbers you haven't checked.
• You think commercially. Margin, risk-adjusted return, payback. You can defend a decision that trades growth for quality.
• You work with AI properly. Real work through Claude Code or Cursor, your own agent skills, and a clear sense of where these tools get things wrong. Agent output is a draft you check.
• You are direct. You disagree with evidence and commit once the call is made.
• You are technical enough. APIs, system design, and a sense of what's actually cheap to build.

Useful, not required:

• Lending, credit risk, collections, or payments
• B2C mobile at scale in an emerging market
• Working inside an engineering team rather than next to one

How we work:

Weekly cycles. Priorities picked Thursday, sprint starts Sunday. Every fourth cycle is for innovation.

No long-term roadmap. We work against corporate goals and change course when the data says to. Prioritization is an argument between product, engineering, and risk — not a scoring spreadsheet.

If you want a stable process and a two-year plan, this will not suit you.