Hello, I'm Emmanuel · open to product roles
I'm a Product Manager with 5+ years leading cross-functional teams to build and scale products across fintech, fraud & compliance, marketplaces and AI, from first user interview to live pilots, paying customers and ₦68M+ in marketplace GMV.
06 products · keep scrolling
Fintech · Fraud prevention
Real-time fraud detection and AML for banks, fintechs and lenders in Kenya and Nigeria. I built the MVP: fraud-scoring rules, SIM-swap signals, AML checks and monitoring workflows that answer allow, review or block and say why. It's now live in a pilot.
RegTech · AML
Fraud detection, transaction monitoring and AML for African banks and fintechs. I defined and built the MVP, from device intelligence and fraud scoring to the monitoring dashboard and automated suspicious activity reports, now processing thousands of transactions in live pilots.
Climate · B2B SaaS
A solar marketplace linking Nigerian homes and businesses with vetted installers. I led the end-to-end build: energy assessment, AI-assisted system sizing, quotes and equipment payments, plus shareable calculator links that turn installers' customers into qualified leads.
AI · Coaching
Founders rehearse a pitch, Podily analyses the delivery and gives real-time feedback and personalised practice, so good coaching isn't limited to people who can afford a coach.
Fintech · Food
RunAm started as a delivery hub for customers, restaurants and riders. The data showed 86% of orders were coming from restaurants, so we pivoted to help food businesses receive orders and payments, and run riders efficiently.
SaaS · Operations
At Atains, a studio building web products and business tools for small businesses, I built a smart inventory app for shops stuck with manual stock counts, overstocking and stockouts, then iterated on the UI from user feedback until satisfaction rose 80%.
I'm a product manager with 5+ years building products for banks, fintechs, marketplaces and small businesses. Since 2024 I've focused on fraud, risk and compliance, where a product decision is the difference between a customer keeping their money or losing it.
I like the zero-to-one stage: talking to users until the problem is sharp, cutting scope to an MVP that proves something, and shipping it with engineering, design and compliance in the same room. I've done that six times, from a food delivery hub to fraud infrastructure now running in live pilots.
I'm technical enough to read the code and prototype with AI tools like Cursor and Claude Code, which makes me faster at turning a problem into something users can try.
A lean, startup mindset: learn fast, cut scope hard, and ship what users will actually pay for.
User interviews, usability sessions and funnel data until the problem is sharp. At RunAm the data showed 86% of orders came from restaurants, and we pivoted.
Podily went from concept to user testing in under three months. Sieve's MVP went straight into a live pilot.
Pricing, monetisation and GTM: Inverta's ₦9,999 subscription plus commissions, ProDetect's pricing tiers and onboarding playbooks.
A shared engineering–compliance process at ProDetect cut discovery-to-delivery time by 25%.
Product strategy · Roadmapping · Monetisation · Activation & retention · Marketplace growth · GTM · Prioritisation
Product analytics · Funnel optimisation · KPIs & OKRs · SQL · Google Analytics · Mixpanel · User research · Usability testing
APIs · B2B SaaS · Marketplaces · Fraud & AML workflows · KYC · PRDs · Agile/Scrum · Python/Django · React · PostgreSQL
AI product integration · LLM agents · RAG · MCP · Evals · Cursor · Claude Code · ChatGPT · Journey mapping · Figma prototyping
Cross-functional leadership · Stakeholder management · Product documentation · Jira · Asana · Trello
I build AI into products where it does real work, and I know where it shouldn't decide. Here's what I use, and where I've shipped it.
Claude and OpenAI models for analysis, summaries and drafting, with Whisper for speech-to-text.
Shipped in Podily · Sieve · InvertaAgents that investigate a case, pull the evidence and recommend the next step, with a person making the call.
Shipped in Sieve · ProDetectGrounding answers in the customer's own data with embeddings and vector search, so the AI cites records instead of guessing.
Similar-case search in ProDetectModel Context Protocol servers that give AI assistants safe, scoped access to product data and actions.
Tools & integrationsTest sets and accuracy tracking before and after launch, so "it seems better" becomes a number.
Sieve: 515-case benchmark, 95% live accuracyKeeping personal data out of prompts, ignoring instructions hidden in records, and AI that recommends but never acts alone.
Sieve Agent · ProDetect CopilotOpen models like Llama and Qwen on Ollama or vLLM for customers whose data can't leave the building.
Sieve on-premiseRubrics, structured outputs and formats people can act on. Podily went from raw JSON to short, scannable feedback.
Podily's 3CTES scoringPrototyping and shipping faster with Cursor, Claude Code and ChatGPT, so users can try an idea in days, not sprints.
Every project since 2024Built product by product
“Quote goes here.”
I'm looking for my next product role in fintech, fraud and risk, marketplaces or AI. If you're hiring, I'd like to hear about it.
