Tobi Okusanya
London,UK
Summary
Data Scientist with 7+ years’ experience applying advanced analytics to build AI solutions that deliver product-level insight and measurable impact. Experienced in experimentation, causal inference and product health metrics covering activation, engagement, retention and LTV. Skilled in deploying scalable ML and experimentation pipelines across BigQuery and Databricks, including sensor-style time-series and cohort retention analysis. Trusted by senior leaders to shape product strategy, improve user experience and drive data-informed decisions.
Overview
8
8
years of professional experience
4
4
years of post-secondary education
Work history
AI Engineer - Client Engineering
IBM
10.2023 - 12.2025
- Automated the ingestion and validation of engagement-event and device-sensor data, creating reliable pipelines for retention and cohort-analysis dashboards. Improved metric freshness from 84% to 99%, enabling faster iteration on product experiments and growth decisions.
- Partnered with product, compliance and engineering teams to develop an ML-driven transaction review system for a fintech client. Reduced false positives by 27%, enhanced fraud detection precision and influenced product roadmap decisions through data storytelling and Tableau insights.
- Built a scalable forecasting pipeline using Databricks, collaborating with data engineers to enhance data ingestion and ensure high integrity of institutional asset feeds. Improved forecast accuracy by 18% and maintained
- Designed and deployed a Generative AI chatbot powered by LLMs with a hybrid recommender system to deliver personalised product insights and automate customer interactions. Built on GCP with integrated CI/CD and monitoring pipelines; solution now being scaled across three international markets hence improving self-serve resolution and increasing feature adoption.
- Developed modular dbt models orchestrated with Airflow to transform raw customer and transaction data for analytics and ML feature generation, cutting query runtimes by 33% and standardising key business metrics that underpin A/B readouts and product-health dashboards.
Advisory Data Scientist - Client Engineering
IBM
09.2021 - 10.2023
- Spearheaded the design and deployment of stochastic simulation models to evaluate customer behaviour and retention scenarios, enabling the team to quantify uncertainty in engagement and churn forecast - hence reducing projected churn by 15%.
- Led a team to create a custom NLP pipeline combining speech-to-text, sentiment scoring and topic modelling for a leading UK bank. Analysis uncovered a 21% link between negative call sentiment and churn, driving targeted retention initiatives and improving customer experience.
- Optimised feature rollout and resource allocation for a global fintech client by building linear programming models, improving delivery efficiency by 15% and enabling data-driven prioritisation of product experiments.
- Architected unified data models integrating customer, transaction and product data to power personalised investment recommendations and reporting. Reduced metric discrepancies by 41% and improved dashboard accuracy across Looker and Power BI, enabling consistent insights and clearer data storytelling across teams.
- Delivered simulation-based optimisation for capacity planning under stochastic demand thus reducing bottlenecks by 20% and improving throughput predictability. Deployed on GCP with CI/CD.
Senior Data Analyst
Ted Baker
10.2020 - 09.2021
- Evaluated regional marketing performance using marketing mix modelling to isolate causal drivers of sales and ROI. Uncovered a strategy that lifted sales volume by 11% and informed budget reallocation across additional UK regions.
- Analysed shopper behaviour and segmentation patterns to identify ‘peak-only’ customers and model retention likelihood. Insights informed CRM experimentation and campaign timing strategies during high-traffic retail periods.
- Automated executive-level reporting pipelines in Python, reducing delivery time by 33% hence increasing data accessibility across finance, digital and merchandising teams.
Data Scientist
XLN Telecom
04.2019 - 09.2020
- Built a churn-prediction model to surface at-risk users and test targeted win-back strategies through A/B experiments, leading to higher customer retention.
- Developed a scalable tariff-fairness scoring model tailored by tenure and history, rolled out across all B2C customers leading to a 17% reduction in customer complaints.
- Devised ARIMA forecasting for pricing transitions that saved approximately £5M over 4 years and informed long-term pricing strategy.
Junior Data Scientist
Pairview
05.2018 - 03.2019
- Supported the development of a churn prediction model with 89% precision and applied association rule mining to uncover cross-sell opportunities, saving £350K annually.
Education
Integrated MEng (Hons) - Chemical Engineering
University of Sheffield
01.2013 - 01.2017
Skills
- Languages & Tools: Python, SQL, dbt, Git, PySpark, Power BI, Tableau, Airflow, Adobe Analytics
- Data Platforms: BigQuery, Databricks, Snowflake, GCP, AWS, IBM Cloud
INTERESTS
Basketball Coach, Bouldering, Competitive Chess, Coding Mentor, Technical Presenter
Timeline
AI Engineer - Client Engineering
IBM
10.2023 - 12.2025
Advisory Data Scientist - Client Engineering
IBM
09.2021 - 10.2023
Senior Data Analyst
Ted Baker
10.2020 - 09.2021
Data Scientist
XLN Telecom
04.2019 - 09.2020
Junior Data Scientist
Pairview
05.2018 - 03.2019
Integrated MEng (Hons) - Chemical Engineering
University of Sheffield
01.2013 - 01.2017