Majid Al Futtaim x I.AM+ (will.i.am) · Lead Product Manager and Orchestrator · 2020 to 2021
Transformed a stalled concept into a production-tested AI super app within six months.
Assumed ownership of a year-old MAF and I.AM+ AI super app concept, repositioned it from competitor to channel, streamlined over 30 features into a viable MVP, and delivered a 14-service, production-tested platform in six months without formal authority.
Executive Snapshot
- MAF and I.AM+ announced a regional AI super app at Davos 2019. Over a year later, no product had shipped and the initiative faced cancellation.
- The MVP expanded to over 30 features. Business units resisted due to channel conflict concerns, and teams across four continents lacked a unified delivery model.
- I was brought in to manage the MAF side of the partnership and assumed ownership of the product blueprint, integration model, and operating layers.
- I streamlined the feature set to a focused MVP, repositioned IAMAI as a channel for business units, and secured cooperation by building relationships with each business unit lead.
- Within six months, IAMAI progressed from concept to production testing across all 14 services, with a complete operating layer and internal beta. The program did not reach public launch due to partnership and leadership changes in the summer of 2021.
The Challenge
MAF and I.AM+ announced their partnership at the World Economic Forum in January 2019. When I joined in April 2020, the program had strong executive sponsorship but remained over a year past its announcement with no shipped product.
The MVP expanded to over 30 features, making prioritization unmanageable. Business units at Carrefour, VOX, and SHARE were concerned IAMAI would compete with their own apps and customer relationships. The I.AM+ engineering team operated across Los Angeles, London, Singapore, and India. I managed the MAF relationship from Dubai without formal authority over any required teams.
The Arabic AI layer needed to support multiple dialects on early-generation LLM models, with robust grounding and guardrails to prevent fabrication. MAF's security and data teams imposed strict requirements. Delays had already occurred, and further stalling risked project cancellation.
My Mandate
- Full product ownership on the MAF side: blueprint, integration model, and all operating layers.
- Reducing the 30+ feature wishlist to a viable MVP and holding that scope against stakeholder pressure.
- Rebuilding business unit relationships at Carrefour, VOX, and SHARE, and turning resistance into active participation.
- Designing every integration layer: product, partner, loyalty, payment, AI, admin, customer support, and data governance.
- Leading delivery without formal authority over any team in the structure.
What I Changed
Reduced scope to a guiding principle rather than a feature list.
I narrowed over 30 features to 14 services, shifting scope management from negotiation to a shared standard. Once the MVP was defined, all out-of-scope requests received principled, consistent responses.
Repositioned IAMAI as a channel rather than a competitor.
Business units initially resisted due to channel conflict concerns. I invested four months in individual relationship-building, reframing the app as a tool to drive customers to existing brands. This approach converted key internal opponents into active supporters.
Validated market need before scaling.
With no evidence of customer demand, I led customer interviews, focus groups, and market research before committing to the full build. This provided executive leadership with an evidence-based investment case and grounded team decisions in real market signals.
Established Arabic guardrails from day one.
With early-generation GPT-3 models and multiple UAE dialects, fabrication was the primary AI risk. I prioritized Khaleeji, Egyptian, and Levantine dialects and implemented a grounding layer to restrict responses to verified product and service data.
Key Decisions and Tradeoffs
Cut to 14 services and hold the line.
Every feature request beyond the agreed MVP was deferred to the post-launch roadmap. The discipline of a shared MVP definition, rather than individual judgment, enabled effective scope governance across more than 20 stakeholders.
Tradeoff: Some features for influential stakeholders were delayed.
Executive signal: I separated scope decisions from politics and established the MVP definition as a shared principle.
Invest four months in alignment before building.
The build began in August 2020, four months after I joined. This period was dedicated to stakeholder alignment, relationship-building, and market validation. Without this groundwork, the build would have lacked a stable foundation.
Tradeoff: The start was delayed for a program already over a year behind.
Executive signal: I identified the root cause as organizational and addressed it as the primary delivery risk.
Dialect prioritization over coverage breadth.
Supporting all Arabic dialects on GPT-3 was not feasible. I prioritized the three dominant UAE dialects and developed a guardrail layer to ensure answer reliability, accepting narrower coverage for more trustworthy outputs.
Tradeoff: Reduced dialect coverage for grounded, reliable responses.
Executive signal: I prioritized user trust over feature completeness in an Arabic-first product.
Results and Validation
Delivery outcomes
- Concept to production testing in six months
- 14 integrated services under one account, wallet, loyalty system, and AI engine
- Internal beta with hundreds of MAF testers
Platform outcomes
- Real-time payment, refund, and partner reconciliation across all services
- Complete operating layer: admin dashboard, partner platform, SOPs, reconciliation, security, data governance, and loyalty extension framework
Strategic outcome
The platform demonstrated the viability of an Arabic-first conversational super app architecture, though partnership and leadership changes ended the initiative before public launch.
The program did not proceed to public launch. Partnership and leadership changes ended the initiative in summer 2021, as publicly reported. The platform reached production and demonstrated the architecture's viability. This is not a market-traction story. It is a production-readiness and transformation-leadership case: reframing strategy, aligning resistant stakeholders, reducing scope, building an Arabic-first AI operating model, and advancing a stalled board-level initiative to production testing.
Press Releases
- Majid Al Futtaim Official Press Release
- Zawya: Majid Al Futtaim and IAM+ join forces to introduce groundbreaking AI solutions
- Forbes Middle East: MAF partners with will.i.am's AI firm
- The National: Hundreds of millions spent on AI as will.i.am tie-up continues
- Images Retail ME: MAF to personalise shopping experience with I.AM+
What This Proves
This case demonstrates the ability to take a stalled, complex GCC initiative with no unified ownership, reframe it organizationally, build cross-functional alignment without formal authority, and deliver a multi-service production platform in six months. The combination of scope governance, stakeholder management across C-suite and business units, Arabic-first AI design, and multi-layer technical delivery is directly applicable to senior digital or AI transformation mandates involving family-group or joint-venture complexity.
Lessons Learned
- Organizational alignment is the first product delivery decision. No level of technical quality can compensate for a misaligned stakeholder model.
- MVP scope governance requires a guiding principle, not a feature list. Once the MVP is defined as a shared standard, scope decisions become objective rather than political.
- Arabic AI requires intentional design. Dialect prioritization, grounding layers, and fabrication prevention must be addressed as primary architecture decisions, not as afterthoughts.