Tijani Matallah

- Built a serverless AI recommendation engine (AWS Lambda, Aurora, S3, Bedrock) handling 900k+ multilingual products in real time. - Implemented semantic search with pgvector + Bedrock and event-driven workflows (SQS + Lambda), cutting participant matching from hours to <5s.

Designed and shipped internal AI & automation tools used by 50+ Partner Program Specialists across EMEA to reduce SOP search time and accelerate decision-making across global partner programs. • Built and launched an AI Knowledge Assistant (Amazon Q Business) reducing SOP lookup time by ~50% and cutting average case resolution from 15min to <5min • Led the end-to-end design of a real-time documentation pipeline (S3 → Redshift → QuickSight), eliminating 100% of manual updates and enabling 3× faster reporting on partner operations • Partnered with program leads to define requirements and deliver automated validation tooling, saving 150+ hours/quarter through scripted workflows adopted across multiple GEOs • Designed event-driven cloud architectures (Lambda, S3, EventBridge) ensuring >95% accuracy in policy updates across 4+ global partner programs (ACE, SPO, SPP, MAP)

Worked on internal product initiatives improving catalog intelligence and user experience across Europe. Developed Gen AI tooling (Amazon Bedrock) and data-driven strategies reducing product duplication and guiding shopping behavior at scale. • Identified a €50M/Y loss from catalog redundancy and led the solution design of a GenAI classification tool (700+ lines Python) • Improved catalog CX and search relevance by consolidating duplicate products and redirecting traffic to multipacks • Partnered with suppliers (BIC, Exacompta, ACCO) and internal stakeholders to deploy product changes across EU marketplace • Conducted business impact analysis to measure profitability improvements and support strategy adoption company-wide