Nabaa Refai
About Career Projects Get in touch
Abu Dhabi, UAE · AI Product & Data Leader

I build AI products that deliver measured outcomes.

12+ years shipping data and AI products across consulting, pharma, travel and an AI SaaS start-up: ML models that saved £2m in year one, a platform that cut a 40-hour task to 15 minutes. I lead teams, and I still own the technical decisions.

Get in touch LinkedIn ↗
£2m+savings in year one from an ML optimisation product at TUI
~99%reduction in proposal drafting time on FundRobin (40 hrs → 15 min)
$250k→$1.5mannual recharges, scaling AstraZeneca's process mining capability
5→20team growth, hired and led across Europe, USA and APAC

From consultant to founder.

01 — About

I started in operational research at EY, building optimisation and simulation models for banks, airlines and major institutions. At TUI I moved from advising to shipping — productising an ML solution for hotel bed allocation that saved ~£2m in its first year.

At AstraZeneca I joined the process mining team as one of its first members and helped grow it from an emerging capability into an embedded, enterprise-wide Centre of Excellence — a $1.5m annual recharge business with a team of 20, delivering measurable outcomes like 47% faster lead times in Clinical Trials Operations.

Now I'm co-founder and Chief Data Officer of FundRobin, an AI grant management platform. I built the RAG pipelines and LLM features myself, and I own the product roadmap. The common thread across all of it: work out what AI adoption actually looks like for an organisation, then build it.

Based in
Abu Dhabi, UAE
Experience
12+ years
Sectors
Pharma · Travel · SaaS · Consulting
Education
BSc Mathematics, Operational Research, Statistics & Economics — University of Warwick, United Kingdom · 1st Class Honours
Speaking
Celonis conferences
Currently
Co-founder & CDO, FundRobin

Where I've worked.

02 — Experience

FundRobin

Co-founder & Chief Data Officer 2025 — PRESENT · ABU DHABI

AI grant management SaaS for nonprofits, charities and universities.

  • Built Smart Matching, Smart Proposal and Robin AI Assistant (OpenAI + Vertex AI LLMs) — cutting proposal first-draft time from 20–40 hours to 15 minutes.
  • Designed the end-to-end pipeline (scraping → chunking → embeddings → retrieval), indexing 4,000+ opportunities from 250+ donors.
  • Python/FastAPI backend on GCP with CI/CD via GitHub Actions; full startup ownership alongside the CEO: roadmap, go-to-market, customer acquisition.
~99% faster drafting4,000+ opportunities250+ donors

AstraZeneca

Director, Data & Advanced Process Analytics 2020 — 2024 · CAMBRIDGE, UK

Helped grow the process intelligence capability from an emerging function into an embedded, enterprise-wide Centre of Excellence within Global Business Services.

  • Joined as one of the team's first members; helped scale the capability from $250k to $1.5m annual recharges and the team from 5 to 20 across Europe, USA and APAC.
  • Expanded from Finance into Operations, Procurement, R&D and Commercial: 10+ processes, 13+ systems, 150+ users, 10M+ records.
  • Delivered 47% faster lead times and 2,000+ hours saved through automation in Clinical Trials Operations; supported the SAP ECC → S/4HANA transformation.
$250k → $1.5mTeam 5 → 2047% faster lead times

TUI

Manager, Data Science & Enterprise Analytics 2017 — 2019 · LONDON, UK

Analytics and data science delivery focused on adoption and commercial outcomes.

  • Built and productionised a Python ML optimisation solution for hotel bed allocation — ~£2m savings in year one, acting as Data Product Owner through rollout.
  • Scaled the analytics team from 4 to 12; secured executive sponsorship through discovery workshops and business-facing demos.
~£2m saved in year 1Team 4 → 12

EY

Senior Consultant, Operational Research & Business Modelling 2013 — 2017 · LONDON, UK

Optimisation, analytics and simulation for clients across banking, aviation, energy, government and telecoms.

  • Built a retail network optimisation solution for a Tier 1 investment bank using simulated annealing — enabling scenario testing for branch strategy; managed a team of three and trained the client team on the model.
  • Designed a bespoke agent-based crowd simulation for a major London institution, informing decisions from staffing through to structural alterations to the building.
  • Investigated performance drivers for a major UK airline with statistical analysis and dynamic visualisation (R, Spotfire).
  • Member of EY's Innovation team, embedding AI and ML into the firm's internal processes; delivered training across the firm globally.
5 industriesGlobal training delivery

Things I've shipped.

03 — Projects
AI SaaS · Co-founder

FundRobin

AI-powered grant management platform. Smart Matching surfaces funding opportunities organisations didn't know existed; Smart Proposal drafts in 15 minutes what used to take 40 hours. Robin AI Assistant answers questions over a RAG pipeline backed by a vector database.

Python · FastAPI · GCP · Vertex AI · OpenAI · Vector DB
fundrobin.com ↗
Side project · Property data

PropIntel

Price per square foot is the most useful metric for comparing UK property values — but no free tool calculated it. I joined HM Land Registry price data with EPC floor areas via a fuzzy address-matching pipeline, and built the stats a property investor actually needs.

Async Python · FastAPI · SPARQL · React · Mapbox
propintel.uk ↗

What I bring.

04 — Capabilities

AI product leadership

Taking AI from idea to real-world use — shaping what to build, delivering it, and embedding it into day-to-day operations. LLMs, RAG pipelines, NLP, ML optimisation.

Data & AI engineering

Python, FastAPI, SQL, GCP (Vertex AI, BigQuery, Cloud Run), vector databases, Snowflake, Supabase, Docker, GitHub Actions. All of it hands-on and in production.

Process intelligence

Enterprise process mining at scale — Celonis (advanced PQL), process transparency for major transformations like SAP ECC → S/4HANA, and turning insights into automation.

Leadership & execution

Building and scaling teams (5→20, 4→12), winning senior sponsorship, and delivering programmes end-to-end.

Cross-industry problem-solving

A consulting foundation, then banking, aviation, energy, government, travel, pharma and SaaS. I'm naturally curious, so new domains and unfamiliar problems don't faze me.

The technical–stakeholder bridge

Equally at home in the codebase and in front of senior stakeholders. Most of my value sits in translating between the two.

Get in touch.

06 — Contact

I'm based in Abu Dhabi and open to data & AI leadership roles where I can build as well as lead, plus advisory work. Email or LinkedIn is the fastest way to reach me.

nabaa.refai@gmail.com LinkedIn ↗
© Nabaa Refai Abu Dhabi · London