Transforming Food Alerts for the FSA

Transforming Food Alerts for the FSA

We began working with the Food Standards Agency back in 2017, transforming the way their data could be handled and used, turning unstructured data into useful information that could be easily used as open source data by other government departments, NGOs and private organisations.

One of the FSA’s key roles is responding to concerns about food safety. Having up to a couple of hundred hits a year, food alerts took a considerable amount of staff time to manage. The data could not be used easily beyond alerting the end user to an issue, despite having a much wider range of potential applications. The largely manual process of publishing an alert, wasn’t built to scale well or cope with a greater number of alerts going through it.

With a huge increase in the number of food manufacturing businesses in the UK over the last few years, the FSA now handles many more alerts a week. Despite the increase in demand on the service, it's able to publish these efficiently, because of the improvements to the process and structure behind the system.

Transforming Food Alerts for the FSA Logo

Client: Food Standards Agency
Our Role: Creating structured data

Digital transformation

As part of the digital transformation program, one of the biggest projects was to streamline the food and allergy alerts (product recalls) that both manufacturers and consumers rely on.

Kicking off the project

The FSA recognised that the information within the food alerts couldn’t be easily used by other organisations and internally, the process was manual and slow.

The comms team would first receive a food alert as a word document. They would then begin the task of copying it across to their website, often having to modify their template where the information didn’t fit. They’d write copy to be used in an SMS and email, before sending it out to relevant parties. It was a half hour process for each food alert, some of which would be received out-of-hours by staff, not necessarily used to working with the data.

None of the data on the FSA published was structured, so anybody wishing to make use of it—NGOs, governmental departments and private companies—would have had to have scraped the website. In reality, this meant that the data was rarely shared or used effectively. This was the main driving factor of the project: improve the digital processes to make the information more accessible and usable.

Creating structured data

Our team created a system whereby incoming alerts are structured, which allows each piece of data to automatically populate the external comms with the relevant information. The ‘tagged’ information is pulled into a template for the FSA website and it generates a shortened URL, useful for sharing on multiple platforms. It also creates the text for an alert email and SMS for anyone signed up for those services, and a summary ready for manually releasing on Twitter.

A welcome by-product of the improvements is the speed at which alerts are generated and shared. This has allowed the FSA to become more efficient, as staff are freed up for other tasks.

As structured data, users can now search for a specific allergen and get up-to-date information about product recalls relating to that foodstuff. The information on the website no longer needs to be scraped to be useful. It is readily available to all who need it through an API. Without manually having to publish them, food alerts now reach the rest of the supply chain and consumers minutes after the FSA receives them. Products can now be recalled at a faster rate, improving safety standards across the food sector.

Since the process has improved, so too has the way it is being used. Largely due to enhanced visibility, the FSA have received a dramatically increased number of food alerts, suggesting that companies are finding it easier to be proactive.

The FSA’s own scientists have themselves been able to take advantage of the new system. It’s much more straightforward to analyse the data, given that all information is now tagged automatically.

Project Summary

Realising value from open source data

Published as open source data, private companies can make commercial use of it by integrating it into their own software, or analysing the data, adding value to their customers.

agileChilli, a Bristol based software company for food manufacturers pulls in the FSA alerts via the API into their NPD software. By cross checking the suppliers and ingredients mentioned in each alert with their own information, the software can flag particular alerts that may need action. Providing all their suppliers and ingredients used are in the software, this saves manually having to check each alert. Analysis of the alerts is also possible within agileChilli software to enable you to streamline your supply chain .


"The Epimorphics API was really well designed, which made it easy to integrate into our workflow using standard tools such as Zapier. It's been running for a couple of years now and has been robust and reliable."

Oliver Kohll, CTO, agileChilli

More Case Studies

Land Registry

Land Registry

Epimorphics supported Land Registry in building capability and supporting the publication of the UK House Price Index as open linked data. The Land Registry records the sale of residential and commercial properties and this data is made available to support many types of users including the public, companies, institutions,...

Read more
Natural Resources Wales

Natural Resources Wales

The Natural Resources Wales Bathing Water Quality Data service was developed with the specific needs of Wales and dual language capabilities built in. The NRW service and the data API all are able to respond to language requests supporting both NRW’s and other Wales user needs

Read more
BBC Sport

BBC Sport

We worked with the BBC to enhance the coverage of major sporting events by introducing linked data. They have gone on to publish the structure that was developed, and use the ontology across the BBC to enhance the experience for website users.

Read more

If you would like to see how we can help structure your data and make it more usable get in touch with us at