Senior Data Science & AI Leader
Dubai, United Arab Emirates
I build AI systems that go live, replacing slow manual processes with data science and automation inside large enterprises.
01 — At a glance
15+
Years of experience in data science, machine learning and AI
$6B USD
Procurement spend base covered by my AI and digitisation programmes, across 20+ countries
€12M
Annual savings identified through machine learning, deep learning and NLP price optimisation
Top 0.1%
Vodafone Top Talent 2021 and 2022, and a top 0.1% performer in FY 2021-22
~$1M USD
Year-one impact delivered in my current role through AI, GenAI automation, vendor savings and revenue uplift
02 — Profile
"Most AI projects do not fail because of the technology. They fail because they never make it into real, everyday use." Closing that gap has been my work for fifteen years.
Over 15+ years I have built AI systems that do not stay as pilots. They go live, replace the process that was there before, and start delivering value early.
My path has run through telecom and procurement — an unusual pairing that has let me see AI from both ends: network-level systems on one side, large-scale business operations on the other.
Recently I have led AI across a global procurement organisation covering 20+ countries. We built platforms in house, replaced existing tools, and rolled out contract intelligence and supplier workflows that now run day-to-day operations. Before that I spent several years in telecom, building AI for network planning, forecasting and customer systems, and leading a team of 15 data scientists across Europe, Africa and APAC.
What I care about most now is where AI is heading next: systems that do not simply assist, but take action and handle real workflows.
If you are working on AI at scale, I am always glad to talk.
03 — Experience
Fifteen years across telecom, software and global procurement, from research teams to leading AI for a $6B USD spend base.
– Present
I lead AI and digitisation across group procurement and a $6B USD spend base spanning 20+ countries, heading a 10+ person delivery structure. My team built proprietary AI spend analytics that replaced a vendor platform, delivered LLM-powered contract intelligence, and reduced sourcing cycle time by 50% — together delivering about $1M USD of year-one impact.
–
I delivered consumer marketing intelligence for millions of customers, covering customer segmentation, campaign intelligence and personalised marketing. I also built e-commerce use cases such as Segment-of-One and Unlimited Free Delivery campaigns, using Python, DataRobot and Oracle SQL.
–
I led a team of 15 data scientists across Europe, Africa, Safaricom and APAC. We built machine learning, deep learning and NLP price optimisation models (CatBoost at 95% adjusted R²) that identified €12M in annual savings across a €13.4B spend base, AI-driven network planning for Safaricom across 8M+ network cells that cut capital expenditure by 10%, and a customer segmentation engine behind Segment of One and Free Delivery personalisation.
–
I developed course material and project work for the upGrad data science programme.
–
Pro bono product work with a startup building autism detection for children, as principal data scientist on key product building.
–
Reporting to the Head of AI, I led a team of 15+ data scientists, machine learning and AI engineers by early 2020. I delivered AI solutions for manufacturing, healthcare and logistics clients, as well as an e-Government innovation platform.
–
Joined as Senior Data Scientist and was later promoted to Chief Data Scientist at this product startup from Moolya. I worked on organisation growth through data-driven decision making and led the development of key products that drove revenue growth.
–
Joined as a Data Specialist and worked across multiple domains on product building and the automation of existing products.
–
04 — Brands & clients
Organisations and platforms I have worked with.
05 — Articles & publications
Frontier Scientific Publishing · 2024
A research paper on applying deep learning to security challenges in intelligent transportation systems, published in the Journal of Artificial Intelligence.
Testing Planet Magazine · 10 April 2011
An article written for Testing Planet Magazine on software testing and games.
06 — Business outcome
Impact
Replaced a commercial vendor platform with proprietary AI spend analytics, saving $500K USD from day one.
Impact
Built a machine learning, deep learning and NLP price optimisation framework across a €13.4B spend base.
Impact
Delivered ContractsGPT, an LLM-powered contract intelligence platform that processed 5,000+ contracts.
Impact
Reports for the leadership team used to take about 4.5 days. I automated the whole set so it runs on a single click.
Impact
Combined cost savings, revenue uplift, GenAI automation and end-to-end digitisation, delivered within the first 12 months.
Impact
Led the end-to-end digitisation of service revenue processes as a flagship transformation outcome.
Impact
Built a detection model for a logistics partner that generated €2.5M in annual savings (2019).
Impact
Delivered an IoT sensor AI solution for predictive maintenance (2017).
Project
Built AI automation for healthcare and visa processing workflows in 2022, applying workflow automation to regulated operational processes.
Project
Delivered a customer value management and loyalty scoring engine in 2022, supporting personalised engagement and retention.
Project
Developed automated machine-learning bug detection at Appachhi in 2018, applying AI to software quality and testing workflows.
07 — Awards & recognition
Recognised as Vodafone Top Talent in 2021 and 2022; the 2022 award went to the top 0.1% performer in FY 2021-22 for building innovative and impactful products for the telecom provider.
Recognised with Best Practices awards for software quality and AI delivery during my years at Vodafone.
Recognised for the turnaround of complex machine learning projects. Some projects were affected by data protection issues, higher attrition, lower morale, poor requirements and agile ways of working; I got them back on track, with appreciation from leadership and business client teams.
My AI teams won awards such as "Most innovative team in South Asia" by Wealth and Finance International, and were top winners at Megathon events for their innovative ideas and implementation.
Worked with multiple groups that organise AI hackathons: mentor at the HerTech Hackathon in AI by Thoughtworks and Facebook, and judge at the Megathon organised by e-Zest and Fresco Capital.
The financial reports presented to the leadership team were manual, taking about 4.5 days each with frequent delays and errors. I automated all of them: one click, 24 seconds.
08 — Volunteering
– Present
Mentoring startup employees in data science, marketing and business growth.
– Present
Developing an app for second opinions for cancer patients.
– Present
Mentoring five data scientists each year.
– Present
TEDx speaker at Vodafone (May 2021) and speaker at PUG on artificial intelligence and quality.
09 — Certificates
10 — Specialities
B.E. Electronics and Telecom
G H Raisoni College of Engineering
2006–2010
11 — Conversations
Four areas I am always glad to discuss.
Machine learning, deep learning, NLP and large language models — and where they genuinely help.
How an AI roadmap moves past the pilot stage and becomes something the business runs on.
Building AI in-house, replacing vendor platforms, and running it across 20+ countries.
Automating the work that slows teams down, from financial reporting to supplier workflows.
12 — Contact
For a conversation, reach out directly or connect on LinkedIn.