I build the systems behind payments and loyalty products.
7 years turning commercial complexity into shipped product across marketplaces, fintech infrastructure and regional loyalty platforms in Spain, the Netherlands and the GCC. I use AI to build faster, and I build the AI myself.
Dubai, UAE
- HousingAnywhereAccount management2019–20
- ExtiaMarket research2020–21
- Alibaba · MiraviaE-commerce2021–22
- ComerciaPayments infrastructure2022–25
- Orbidi0-to-1 SaaS2025
- Majid Al FuttaimLoyalty & payments2025–now
Every one of those roles had the same shape underneath: a business problem with real money attached, a user who didn't care about our internal complexity, and a small window to prove something worked before asking for more investment. I've spent seven years getting good at that window.
Right now I lead product for a regional co-brand card and loyalty program used by 500k+ monthly active users across 5 GCC and MENA markets, while running an independent AI practice on the side, building the kind of tools I wish I'd had earlier in my career.
What I believe
- Understand the payment or loyalty problem from the user's side before touching the roadmap.
- Ship the smallest version that proves the mechanism, not the polished version.
- Measure with a real control group. A before/after chart is a guess wearing a chart's clothes.
- Use AI to move faster through the boring parts, never to skip the thinking.
Four products, four very different constraints.
Partner and client names below are generalized where confidentiality applies. Every metric is real and something I can walk through in detail.
Context
A regional loyalty and payments platform serving 500k+ monthly active users and billions of AED in annual transaction volume across UAE, KSA, Egypt, Bahrain and Oman. My scope: co-brand credit card acquisition and the loyalty experience for existing cardholders, working directly with regional banking partners.
Discovery
Acquisition drop-off was concentrated at a specific step in the sign-up flow, and existing welcome incentives weren't translating into activated, engaged cardholders. I needed a change I could actually attribute, not a redesign I'd have to take on faith.
Solution
- Redesigned the co-brand card acquisition flow and ran a controlled A/B test across 4 banner variants.
- Launched a new Welcome Bonus initiative with a real-time redemption tracker so cardholders could see their own progress instead of a static promise.
Impact
Illustrative diagram of the redesigned flow and where the A/B test sat. Screens are not shown for confidentiality; the 15% acquisition result is real.
Context
Comercia runs payment infrastructure for merchants across card schemes (Visa, Mastercard, Amex) and local alternative methods. I owned the Pay by Link product and led Plazox, Comercia's own BNPL solution for a network of Spanish banks, while managing a team of 2 Product Managers.
Discovery
Pay by Link usage was healthy on first use but weak on repeat use, which meant merchants were paying for infrastructure their customers weren't returning to. Separately, several enterprise retail and travel clients needed a BNPL option Comercia didn't yet own end to end.
Solution
- Redefined the Pay by Link product flow to reduce friction on repeat transactions.
- Led product for Plazox from concept through launch with the partner bank network.
- Personalized PSP integration and checkout for major retail and travel merchants across card schemes and Pay by Bank methods including Bizum.
Impact
Pay by Link performance before/after the redesign, plus the Plazox BNPL initiative shipped in parallel.
Context
Orbidi is a B2B SaaS marketing tool for SMEs, helping small business owners (many in their 50s and 60s, with no marketing background) generate social content for Instagram and LinkedIn. I owned the roadmap and KPIs and worked directly with the CEO on pricing.
Discovery
Growth had stalled and the team's instinct was to discount harder. I wanted to know which lever actually moved retained users, not just signups, so I ran three approaches in parallel instead of guessing: bundled pricing, a free trial, and a referral program, and measured each one separately.
Solution
- Bundled pricing offers, tested in isolation: a real but modest 4% lift.
- A free trial promotion that looked strong on signups but, once checked against activation, wasn't — see Learnings below.
- A referral program (2 weeks of free premium membership for both sides if the referred user stayed 2+ months), validated with an 80/20 test against a genuine holdout control group, plus a WhatsApp engagement layer that drove plan upgrades.
Impact
Each lever measured in isolation. Only the referral loop, validated against a real holdout group, produced durable growth.
Context
Miravia was Alibaba's new Spain-focused marketplace, built to compete for Spanish sellers and buyers where AliExpress couldn't. I worked across both sides of the marketplace: the seller experience and the late-stage buyer funnel.
Discovery
Sellers were slow to get live because onboarding and product upload weren't built for a Spanish seller base with no experience on Chinese marketplace tooling. On the buyer side, checkout was leaking conversion, and Spanish shoppers expected local payment habits AliExpress never had to support.
Solution
- Built a vendor onboarding flow, including product upload, designed around how Spanish SMEs actually work.
- Integrated Bizum as an alternative payment method directly in the late-stage buyer funnel.
- Led discovery for social commerce inside the core shopping flow, and shipped search personalization and related-search suggestions.
Impact
Two sides of the same marketplace: faster sellers in, smoother checkout and search for buyers.
Things that looked like wins and weren't.
The case studies above show what worked. This is the part most portfolios leave out.
A one-month free trial looked like a clear win: signups rose by roughly 30%. But when I checked what those users actually did inside the product, only about 2% engaged with it at all. The lift was real; the growth wasn't. I killed the initiative instead of reporting the vanity number, and redirected the budget into the referral program that ended up driving the real 15% gain.
What changed after this: I stopped trusting a single top-of-funnel metric on its own, and started pairing every growth experiment with an activation check before calling it a result.
When I split the growth initiatives into separate channels to see what was actually driving results, bundled pricing offers on their own only moved usage by about 4%, far below what the team had hoped going in. It wasn't a failure of the idea, but it told me pricing alone wasn't the lever, and it's why I pushed harder on Plazox and the merchant-specific PSP work instead of iterating further on bundles.
What changed after this: I got much more disciplined about testing one lever at a time before combining them into a single "growth initiative."
AI is my operating layer, not a slide in the deck.
I run an independent AI practice at raquelgalindo.io alongside my full-time role, since December 2025. Below is what I've actually built, not what I've read about.
AI voice reservation assistant
Built for a hospitality client: a phone agent that answers, checks table availability against the reservation system, books the table, and follows up on WhatsApp with a reminder. Delivered in 3 languages (English, Spanish, Catalan).
Consultancy websites
Two production sites (English and Spanish) with an AI savings calculator and a lead-capture chatbot, GDPR/PDPL-compliant, instrumented with Google Analytics and Microsoft Clarity.
Workflow automation
Day-to-day product workflows built with Make and n8n, connecting AI models to Jira, Confluence, Excel and Canva to accelerate discovery, PRDs and documentation, plus an AI marketing agent supporting campaign workflows at Majid Al Futtaim.
This portfolio
Built and instrumented with GA4 event tracking and UTM-based attribution modeling, so I can see which case study actually gets read before a call, not just guess.
AI Tools I Actually Use
Breadth across the PM workflow, not just visual generation. Solid chips are proven daily use; dashed tools are picks I'm actively exploring and folding in.
Swipe to see all columns →
| Workflow stage | Tool(s) used | What it's used for |
|---|---|---|
| Research & synthesis | ClaudeChatGPTPerplexity |
Pulling market and competitive context, synthesizing it into a structured brief before writing a PRD or case study. |
| Writing (PRDs, specs, docs) | ClaudeChatGPTNotion AI |
First-draft PRDs, documentation and case studies, then edited for accuracy and my own voice. |
| Prototyping & building | ClaudeCursorLovable |
Scaffolding functional prototypes and shipping small builds end-to-end, like this portfolio, well past throwaway mockups. |
| Workflow automation | Maken8n |
Connecting AI models to Jira, Confluence and Canva to automate discovery, PRDs and day-to-day product operations. |
| Analytics & reporting | Power BIMetabaseGA4Clarity |
Turning raw product and campaign data into dashboards and plain-English readouts for stakeholders. |
| Voice & conversational AI | OpenAICloudflareTwilio |
Building a phone-based reservation assistant end-to-end, from call handling to WhatsApp follow-up. |
Tools I work in, day to day
Product
- Product Strategy & roadmapping
- Agile & Scrum delivery
- OKR / KPI-driven planning
- Customer discovery & user research
- Go-to-market strategy
Payments & Loyalty
- Card schemes: Visa, Mastercard, Amex
- PSP & Pay by Bank integrations
- BNPL & Pay by Link products
- Co-brand loyalty programs
- A/B testing & controlled experimentation
SaaS & Marketplace
- SaaS pricing & packaging
- Marketplace seller onboarding & ops
- Checkout & payment conversion
- Referral & growth loops
- Search & discovery personalization
Let's talk about what you're building.
Open to Senior Product Manager roles across payments, loyalty and AI-native products, GCC or fully remote.
Tired of reading CVs?
Take a break — help me dodge endless meetings and collect good product instincts instead.
Space, ↑ or tap to jump