Industry
Retail / AI and Data Science
Company
Tesco Technology
Interface
Web
Platform
Internal Enterprise SaaS Platform
Industry
Retail / AI and Data Science
Company
Tesco Technology
Interface
Web
Platform
Internal Enterprise SaaS Platform
Industry
Retail / AI and Data Science
Company
Tesco Technology
Interface
Web
Platform
Internal Enterprise SaaS Platform
Product Overview
Product Overview
Product Overview
Tesco Technology is a product-led tech division of Tesco, the UK's largest retailer, with a 27.8% market share and over £4.3 billion spent on business operations annually. Within its AI and Data Science function, the team develops advanced decision-making tools for complex retail and supply-chain operations. Tesco Technology is widely recognised as a pioneer in UK retail tech innovation. Its Data Science Team's algorithmic work has been awarded the OR Society's President's Medal, alongside numerous other industry accolades.
As a Design Manager, I led the end-to-end redesign of the team's internal business-simulation platform, scaling it from a niche data-science tool into a widely adopted enterprise SaaS product. This internal simulation platform enables business teams and data scientists to model “what-if” scenarios before making real-world operational changes. It supports decision-making across orders, product availability, transport, distribution costs, and supply-chain resilience.
Before I joined the project as a Senior UX designer, the platform had been designed primarily by engineers and data scientists. Although it had strong technical potential, its interface was difficult for business users to understand and use independently. This limited adoption, increased reliance on the Data Science team, and prevented Tesco from fully realising the platform’s commercial value.
Tesco Technology is a product-led tech division of Tesco, the UK's largest retailer, with a 27.8% market share and over £4.3 billion spent on business operations annually. Within its AI and Data Science function, the team develops advanced decision-making tools for complex retail and supply-chain operations. Tesco Technology is widely recognised as a pioneer in UK retail tech innovation. Its Data Science Team's algorithmic work has been awarded the OR Society's President's Medal, alongside numerous other industry accolades.
As a Design Manager, I led the end-to-end redesign of the team's internal business-simulation platform, scaling it from a niche data-science tool into a widely adopted enterprise SaaS product. This internal simulation platform enables business teams and data scientists to model “what-if” scenarios before making real-world operational changes. It supports decision-making across orders, product availability, transport, distribution costs, and supply-chain resilience.
Before I joined the project as a Senior UX designer, the platform had been designed primarily by engineers and data scientists. Although it had strong technical potential, its interface was difficult for business users to understand and use independently. This limited adoption, increased reliance on the Data Science team, and prevented Tesco from fully realising the platform’s commercial value.
Tesco Technology is a product-led tech division of Tesco, the UK's largest retailer, with a 27.8% market share and over £4.3 billion spent on business operations annually. Within its AI and Data Science function, the team develops advanced decision-making tools for complex retail and supply-chain operations. Tesco Technology is widely recognised as a pioneer in UK retail tech innovation. Its Data Science Team's algorithmic work has been awarded the OR Society's President's Medal, alongside numerous other industry accolades.
As a Design Manager, I led the end-to-end redesign of the team's internal business-simulation platform, scaling it from a niche data-science tool into a widely adopted enterprise SaaS product. This internal simulation platform enables business teams and data scientists to model “what-if” scenarios before making real-world operational changes. It supports decision-making across orders, product availability, transport, distribution costs, and supply-chain resilience.
Before I joined the project as a Senior UX designer, the platform had been designed primarily by engineers and data scientists. Although it had strong technical potential, its interface was difficult for business users to understand and use independently. This limited adoption, increased reliance on the Data Science team, and prevented Tesco from fully realising the platform’s commercial value.
What Was the Challenge?
What Was the Challenge?
What Was the Challenge?
Tesco spends more than £4.3 billion annually on business operations. Simulation provides a safe way to test complex operational scenarios before introducing changes that may affect customers, availability, transport, and costs. Previously, when a business unit needed to simulate a scenario, it had to request support from the Data Science team. If the scenario required refinement, the process would repeat. This created long lead times, dependency on a limited pool of specialists, and barriers to faster business decision-making.
Tesco also relied on expensive third-party simulation tools, including GuruX and AnyLogic, costing approximately £100,000 per user each year. These tools were not fully integrated with Tesco’s internal systems, had limited adaptability, and could not easily reflect the complexity of Tesco’s operations.
The internal simulation platform was created to reduce this dependency. However, the original product experience was not designed around business or power users. Its engineer-led MVP interface made the platform difficult to navigate, understand, and use, resulting in very limited adoption. Redesigning and expanding a mission-critical simulation platform with little regard for usability, and turning it into a self-service tool that non-technical business users and data scientists could operate independently, all while replacing costly, poorly integrated third-party software and reshaping how business units across Tesco approached scenario planning and decision-making.
Tesco spends more than £4.3 billion annually on business operations. Simulation provides a safe way to test complex operational scenarios before introducing changes that may affect customers, availability, transport, and costs. Previously, when a business unit needed to simulate a scenario, it had to request support from the Data Science team. If the scenario required refinement, the process would repeat. This created long lead times, dependency on a limited pool of specialists, and barriers to faster business decision-making.
Tesco also relied on expensive third-party simulation tools, including GuruX and AnyLogic, costing approximately £100,000 per user each year. These tools were not fully integrated with Tesco’s internal systems, had limited adaptability, and could not easily reflect the complexity of Tesco’s operations.
The internal simulation platform was created to reduce this dependency. However, the original product experience was not designed around business or power users. Its engineer-led MVP interface made the platform difficult to navigate, understand, and use, resulting in very limited adoption. Redesigning and expanding a mission-critical simulation platform with little regard for usability, and turning it into a self-service tool that non-technical business users and data scientists could operate independently, all while replacing costly, poorly integrated third-party software and reshaping how business units across Tesco approached scenario planning and decision-making.
Tesco spends more than £4.3 billion annually on business operations. Simulation provides a safe way to test complex operational scenarios before introducing changes that may affect customers, availability, transport, and costs. Previously, when a business unit needed to simulate a scenario, it had to request support from the Data Science team. If the scenario required refinement, the process would repeat. This created long lead times, dependency on a limited pool of specialists, and barriers to faster business decision-making.
Tesco also relied on expensive third-party simulation tools, including GuruX and AnyLogic, costing approximately £100,000 per user each year. These tools were not fully integrated with Tesco’s internal systems, had limited adaptability, and could not easily reflect the complexity of Tesco’s operations.
The internal simulation platform was created to reduce this dependency. However, the original product experience was not designed around business or power users. Its engineer-led MVP interface made the platform difficult to navigate, understand, and use, resulting in very limited adoption. Redesigning and expanding a mission-critical simulation platform with little regard for usability, and turning it into a self-service tool that non-technical business users and data scientists could operate independently, all while replacing costly, poorly integrated third-party software and reshaping how business units across Tesco approached scenario planning and decision-making.
Team Contribution and Responsibilities
Team Contribution and Responsibilities
Team Contribution and Responsibilities
As the first designer embedded in Tesco Technology’s Data Science division, I led the end-to-end UX strategy and design for the simulation platform. Working alongside data scientists, backend and frontend engineers, product owners, and business stakeholders, my role covered design strategy and vision, platform architecture, UX/UI design, and end-to-end ownership of the user experience, from research and information architecture through to interface design, usability testing, and delivery alongside engineering, translating complex data-science capabilities into a usable self-service experience and creating a reusable UX framework for AI products aligned with engineering and design standards.
As the first designer embedded in Tesco Technology’s Data Science division, I led the end-to-end UX strategy and design for the simulation platform. Working alongside data scientists, backend and frontend engineers, product owners, and business stakeholders, my role covered design strategy and vision, platform architecture, UX/UI design, and end-to-end ownership of the user experience, from research and information architecture through to interface design, usability testing, and delivery alongside engineering, translating complex data-science capabilities into a usable self-service experience and creating a reusable UX framework for AI products aligned with engineering and design standards.
As the first designer embedded in Tesco Technology’s Data Science division, I led the end-to-end UX strategy and design for the simulation platform. Working alongside data scientists, backend and frontend engineers, product owners, and business stakeholders, my role covered design strategy and vision, platform architecture, UX/UI design, and end-to-end ownership of the user experience, from research and information architecture through to interface design, usability testing, and delivery alongside engineering, translating complex data-science capabilities into a usable self-service experience and creating a reusable UX framework for AI products aligned with engineering and design standards.
Problem Statement
Problem Statement
Problem Statement
General: Tesco previously relied on expensive third-party simulation tools such as GuruX and AnyLogic, costing around £100,000 per user per year. These tools were not integrated with Tesco's internal systems, offered limited functionality, and could not adapt to the complexity of the UK's largest retailer. Business teams required a faster, more independent way to test operational “what-if” scenarios before implementing real-world changes that could affect costs, customer availability, orders, transport, and distribution.
General: Tesco previously relied on expensive third-party simulation tools such as GuruX and AnyLogic, costing around £100,000 per user per year. These tools were not integrated with Tesco's internal systems, offered limited functionality, and could not adapt to the complexity of the UK's largest retailer. Business teams required a faster, more independent way to test operational “what-if” scenarios before implementing real-world changes that could affect costs, customer availability, orders, transport, and distribution.
General: Tesco previously relied on expensive third-party simulation tools such as GuruX and AnyLogic, costing around £100,000 per user per year. These tools were not integrated with Tesco's internal systems, offered limited functionality, and could not adapt to the complexity of the UK's largest retailer. Business teams required a faster, more independent way to test operational “what-if” scenarios before implementing real-world changes that could affect costs, customer availability, orders, transport, and distribution.
Operational: When business units needed to model a scenario, they had to request support from the Data Science team, and any change meant repeating the whole request, creating long lead times and a strong dependency on a small pool of specialists.
Operational: When business units needed to model a scenario, they had to request support from the Data Science team, and any change meant repeating the whole request, creating long lead times and a strong dependency on a small pool of specialists.
Operational: When business units needed to model a scenario, they had to request support from the Data Science team, and any change meant repeating the whole request, creating long lead times and a strong dependency on a small pool of specialists.
Business constraints: The solution needed to support complex real-time simulation data, multiple simulation models, historical comparisons, and integration with Tesco APIs, while remaining understandable and usable for non-specialist business users and data scientists as power users. It also needed to scale across multiple business areas and enable both business teams and data scientists to work more independently.
Business constraints: The solution needed to support complex real-time simulation data, multiple simulation models, historical comparisons, and integration with Tesco APIs, while remaining understandable and usable for non-specialist business users and data scientists as power users. It also needed to scale across multiple business areas and enable both business teams and data scientists to work more independently.
Business constraints: The solution needed to support complex real-time simulation data, multiple simulation models, historical comparisons, and integration with Tesco APIs, while remaining understandable and usable for non-specialist business users and data scientists as power users. It also needed to scale across multiple business areas and enable both business teams and data scientists to work more independently.
Technical: An internal SaaS platform was an MVP engineering solution that was build to remove this dependency, but its interface had been created entirely by engineers with no UX input. Business users found it frustrating to use, which led to low adoption, underutilisation of a platform with real potential, and increased operational risk.
Technical: An internal SaaS platform was an MVP engineering solution that was build to remove this dependency, but its interface had been created entirely by engineers with no UX input. Business users found it frustrating to use, which led to low adoption, underutilisation of a platform with real potential, and increased operational risk.
Technical: An internal SaaS platform was an MVP engineering solution that was build to remove this dependency, but its interface had been created entirely by engineers with no UX input. Business users found it frustrating to use, which led to low adoption, underutilisation of a platform with real potential, and increased operational risk.
Design objective
Design objective
Design objective
Create a scalable, self-service simulation platform that simplifies complex data-science workflows, enables business users to make informed operational decisions independently, reduces dependency on the Data Science team and third-party tools, and supports the future expansion of AI-driven simulation products across Tesco Technology.
Create a scalable, self-service simulation platform that simplifies complex data-science workflows, enables business users to make informed operational decisions independently, reduces dependency on the Data Science team and third-party tools, and supports the future expansion of AI-driven simulation products across Tesco Technology.
Create a scalable, self-service simulation platform that simplifies complex data-science workflows, enables business users to make informed operational decisions independently, reduces dependency on the Data Science team and third-party tools, and supports the future expansion of AI-driven simulation products across Tesco Technology.
Design Process
Design Process
Design Process
1. Briefing
Assessed the platform's technical constraints, existing adoption data, and business priorities across the units it was meant to serve.
1. Briefing
Assessed the platform's technical constraints, existing adoption data, and business priorities across the units it was meant to serve.
Research
Investigated how business users approached planning, the existing engineer-built interface, and how comparable simulation tools structured complex, data-heavy workflows.
Research
Investigated how business users approached planning, the existing engineer-built interface, and how comparable simulation tools structured complex, data-heavy workflows.
Information Architecture
Designed a scalable IA that organised simulation inputs, scenario parameters, and outputs so that non-technical users could configure and run simulations without data-science support.
Information Architecture
Designed a scalable IA that organised simulation inputs, scenario parameters, and outputs so that non-technical users could configure and run simulations without data-science support.
CJM "to Be"
Built a CJM, mapping the shift from a dependency-driven, request-based workflow to a self-service model aligned with business needs and technical constraints.
CJM "to Be"
Built a CJM, mapping the shift from a dependency-driven, request-based workflow to a self-service model aligned with business needs and technical constraints.
Design Concept
Developed a cohesive design concept and the UX framework for AI products, aligning with engineering and data science approach
Design Concept
Developed a cohesive design concept and the UX framework for AI products, aligning with engineering and data science approach
Usability Testing
Led iterative usability testing with business users, refining the interface, UX logic and information architecture after each round.
Usability Testing
Led iterative usability testing with business users, refining the interface, UX logic and information architecture after each round.
Presentation
Presented UX research findings and the finalised design concept to engineering, data science, and business stakeholders to align on scope and implementation.
Presentation
Presented UX research findings and the finalised design concept to engineering, data science, and business stakeholders to align on scope and implementation.
Development
Implemented the approved design consistently across all screens, working closely with UI designer, data scientists and front- and back-end engineers.
Development
Implemented the approved design consistently across all screens, working closely with UI designer, data scientists and front- and back-end engineers.
Launch
Drove the platform through to launch and continued adoption, supporting its rollout across multiple business areas.
Launch
Drove the platform through to launch and continued adoption, supporting its rollout across multiple business areas.
Discovery plan
Discovery plan
Discovery plan
Flow and Information Architecture
Flow and Information Architecture
Flow and Information Architecture
Designed a user-centred structure for the platform that let business units configure "what-if" scenarios, run multi-run and historical-data simulations, and access Tesco's internal APIs, all without needing a data scientist as an intermediary. This included the information architecture for the upcoming Tornado-Hive simulation, built to be scalable and adaptable to future simulation models across Tesco.
Designed a user-centred structure for the platform that let business units configure "what-if" scenarios, run multi-run and historical-data simulations, and access Tesco's internal APIs, all without needing a data scientist as an intermediary. This included the information architecture for the upcoming Tornado-Hive simulation, built to be scalable and adaptable to future simulation models across Tesco.
Designed a user-centred structure for the platform that let business units configure "what-if" scenarios, run multi-run and historical-data simulations, and access Tesco's internal APIs, all without needing a data scientist as an intermediary. This included the information architecture for the upcoming Tornado-Hive simulation, built to be scalable and adaptable to future simulation models across Tesco.
Customer Journey Mapping
Customer Journey Mapping
Customer Journey Mapping
Customer behaviour: Business units had only ever accessed simulations by submitting a request to the Data Science team and waiting for results, a slow, dependency-driven pattern with no self-service option.
Customer behaviour: Business units had only ever accessed simulations by submitting a request to the Data Science team and waiting for results, a slow, dependency-driven pattern with no self-service option.
Customer behaviour: Business units had only ever accessed simulations by submitting a request to the Data Science team and waiting for results, a slow, dependency-driven pattern with no self-service option.
Difficulties: Users had no existing mental model for configuring and running a simulation themselves; the platform's engineer-built interface offered no guidance on how to set parameters, interpret outputs, or compare scenarios.
Difficulties: Users had no existing mental model for configuring and running a simulation themselves; the platform's engineer-built interface offered no guidance on how to set parameters, interpret outputs, or compare scenarios.
Difficulties: Users had no existing mental model for configuring and running a simulation themselves; the platform's engineer-built interface offered no guidance on how to set parameters, interpret outputs, or compare scenarios.
Action: A new journey was designed from scratch, guiding users from scenario setup through to multi-run comparison and historical benchmarking, making self-service simulation intuitive for people with no data-science background.
Action: A new journey was designed from scratch, guiding users from scenario setup through to multi-run comparison and historical benchmarking, making self-service simulation intuitive for people with no data-science background.
Action: It was necessary to create a new behaviour model in the customer journey map in order to anticipate these changes and make it clear to the client what needs to be done to enter the business lounge as part of the premium banking service.
What Was the Solution?
What Was the Solution?
What Was the Solution?
A redesigned, self-service simulation platform that let business units independently model the impact of changes on orders, customer availability, and transport and distribution costs, run multi-run and historical simulations, and connect across Tesco's internal models and APIs. This replaced a costly, dependency-driven process with an in-house tool that also allowed data scientists to build new simulations on the same self-service foundation.
A redesigned, self-service simulation platform that let business units independently model the impact of changes on orders, customer availability, and transport and distribution costs, run multi-run and historical simulations, and connect across Tesco's internal models and APIs. This replaced a costly, dependency-driven process with an in-house tool that also allowed data scientists to build new simulations on the same self-service foundation.
A redesigned, self-service simulation platform that let business units independently model the impact of changes on orders, customer availability, and transport and distribution costs, run multi-run and historical simulations, and connect across Tesco's internal models and APIs. This replaced a costly, dependency-driven process with an in-house tool that also allowed data scientists to build new simulations on the same self-service foundation.
Development
Development
I worked closely with front-end and back-end engineers, data scientists and business stakeholders to ensure that the user experience supported real-time simulation data, access to Tesco APIs, multi-run simulations, historical comparisons, and connected simulation models.
The updated experience allowed users to simulate the impact of operational changes on:
- customer orders
- product availability
- transport and distribution costs
- supply-chain performance
- single and multiple scenario runs
- historical data for comparison.
Quantitative UX KPIs
Quantitative UX KPIs
Quantitative UX KPIs
1. Task success rate
1. Task success rate
1. Task success rate
Improved information architecture and interface design helped business users complete simulation tasks independently with contextual help and hint text.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting in a more streamlined booking process. Leveraging the Jobs to be Done framework and Flow map further refined the user-centric experience.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting
in a more streamlined booking process. Leveraging the Jobs to be Done framework
and Flow map further refined the user-centric experience.
2. Time on task
2. Time on task
2. Time on task
Reduction in average time spent configuring and running simulations, reflecting increased efficiency.
Reduction in average time spent configuring and running simulations, reflecting increased efficiency.
Reduction in average time spent configuring and running simulations, reflecting increased efficiency.
3. Adoption / usage volume
3. Adoption / usage volume
3. Adoption / usage volume
Increase in monthly platform executions, from a handful of simulations to over a thousand.
By integrating attractive design elements and 3D illustrations,
I improved the visual attractiveness and functionality of the reservation and payment process.
Increase in monthly platform executions, from a handful of simulations to over a thousand.
4. User error rate
4. User error rate
4. User error rate
Reduced errors thanks to a clearer, more guided interface built on a tested design framework.
By integrating attractive design elements and 3D illustrations,
I improved the visual attractiveness and functionality of the reservation and payment process.
Reduced errors thanks to a clearer, more guided interface built on a tested design framework.
Qualitative UX KPIs
Qualitative UX KPIs
Qualitative UX KPIs
1. Reported expectations vs performance
1. Reported expectations vs performance
1. Reported expectations vs performance
Business users' expectations of a self-service tool aligned with the actual experience of using the simulation platform, increasing trust in it.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting in a more streamlined booking process. Leveraging the Jobs to be Done framework and Flow map further refined the user-centric experience.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting
in a more streamlined booking process. Leveraging the Jobs to be Done framework
and Flow map further refined the user-centric experience.
2. Overall satisfaction
2. Overall satisfaction
2. Overall satisfaction
Strong satisfaction with the shift from a request-based process to an independent, self-service simulation experience.
Strong satisfaction with the shift from a request-based process to an independent, self-service simulation experience.
Strong satisfaction with the shift from a request-based process to an independent, self-service simulation experience.
How Did I Make an Impact?
How Did I Make an Impact?
How Did I Make an Impact?
1. Enabled Self-Service at Scale
1. Enabled Self-Service at Scale
1. Enabled Self-Service at Scale
Business units went from requesting simulations through the Data Science team to running them independently, removing a key operational bottleneck.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting in a more streamlined booking process. Leveraging the Jobs to be Done framework and Flow map further refined the user-centric experience.
Through in-depth interviews, I gained valuable insights into user needs and motivations, resulting
in a more streamlined booking process. Leveraging the Jobs to be Done framework
and Flow map further refined the user-centric experience.
2. Drove Platform Adoption
2. Drove Platform Adoption
2. Drove Platform Adoption
Monthly platform executions grew from just 7 in 2024 to over 1,300 by February 2025, directly reflecting the impact of the usability and integration improvements.
Monthly platform executions grew from just 7 in 2024 to over 1,300 by February 2025, directly reflecting the impact of the usability and integration improvements.
Monthly platform executions grew from just 7 in 2024 to over 1,300 by February 2025, directly reflecting the impact of the usability and integration improvements.
3. Grew From a Solo Designer to a Team
3. Grew From a Solo Designer to a Team
3. Grew From a Solo Designer to a Team
I joined the project as its first and only designer, covering research, UX/UI design, strategy, and operations single-handedly. Within a year, this contribution led to my promotion to Design Manager, and I now lead a growing team of a Senior UI Designer, Senior UX Designer, Senior UX Researcher, Mid-level Researcher, and Service Designers. What began as a single project has expanded to two, with the simulation platform set to become part of an integrated 0-to-1 platform — a new AI-powered system we're building to help internal business teams secure Tesco's supply chain.
By integrating attractive design elements and 3D illustrations,
I improved the visual attractiveness and functionality of the reservation and payment process.
I joined the project as its first and only designer, covering research, UX/UI design, strategy, and operations single-handedly. Within a year, this contribution led to my promotion to Design Manager, and I now lead a growing team of a Senior UI Designer, Senior UX Designer, Senior UX Researcher, Mid-level Researcher, and Service Designers. What began as a single project has expanded to two, with the simulation platform set to become part of an integrated 0-to-1 platform — a new AI-powered system we're building to help internal business teams secure Tesco's supply chain.
What Was the Final Result?
What Was the Final Result?
The redesign and relaunch of the simulation platform transformed it from an underused, engineer-built MVP into a widely adopted enterprise SaaS platform. The platform now delivers measurable financial impact, with estimated savings and will be part of a new AI platform as part of a security of supply project, while removing Tesco's reliance on third-party tools costing roughly £100,000 per user per year. The project's success led to my selection to lead UX for Security of Supply Chain, Tesco's largest and most strategically important AI initiative.
The redesign and relaunch of the simulation platform transformed it from an underused, engineer-built MVP into a widely adopted enterprise SaaS platform. The platform now delivers measurable financial impact, with estimated savings and will be part of a new AI platform as part of a security of supply project, while removing Tesco's reliance on third-party tools costing roughly £100,000 per user per year. The project's success led to my selection to lead UX for Security of Supply Chain, Tesco's largest and most strategically important AI initiative.
