Following best practices in agile software development to meet actual user needs
Design Methodology
Following best practices in agile software development and user-centered design, the ECHOLOT system will be developed iteratively using the Double Diamond Model, developed by Design Council UK and employed worldwide across public institutions, universities and large technology and product design companies. This model helps to understand challenging problems and develop innovative solutions through four iterative phases. Discovery and Definition phases involve in-depth engagement with stakeholder communities to gather insights. Once the problem(s) are (re)defined in an innovative way, the model works through the Implementation and Evaluation phases. Possible solutions are evaluated through test implementations, and are improved incrementally. Our methodology also involves co-design tools such as user personas, scenarios and journey maps.
User Personas
A user persona is a realistic, research-based profile of a target user, summairising their goals, needs, behaviors, and pain points—it is a tool we use to design with a clear, shared understanding of who we’re building ECHOLOT for. The user personas for ECHOLOT were developed together with our case study stakeholders. Each persona shares goals, needs and pain points which form the backbone of ECHOLOT’s requirements engineering process.
Goals: Nina wants to increase data findability and better support cross-departmental work in her institution. Enhancing existing collection data with external resources is also important for presentation purposes (e.g. in museum displays, not just back-of-house). She aims to follow up data exports to Wikidata with editorial work on Wikipedia.
Needs: Nina needs guidance on ontology modeling and practical advice on data enrichment. She would like to work with semi-automated data entry workflows and use standard schemas. She also needs statistics on how collection data is being accessed and used.
Pain points: Nina is held back by the manual effort required by current data curation processes and the lack of automation. Further challenges are posed by limited staff and resources, skills gaps, and inflexible legacy systems.
The Museum Curator
Goals: Andreas wants to share collection data seamlessly across multiple platforms and data aggregators, and link to authority files where relevant. A core ambition is reliable roundtripping – being able to publish data and pull back crowdsourced erichments without losing integrity.
Needs: Andreas needs structured workflows for data deduplication and transformation, terminology mapping and entity linking. Data validation is also a requirement. Automating the alignment of local entities to Wikidata IDs is a particular priority for the roundtripping process.
Pain points: Andreas struggles with the limitations of tools like OpenRefine, which don’t scale well for his workflows. The absence of real-time data sync with some platforms adds friction, while data silos make it hard to get a unified view. Incomplete records, messy data (e.g. inconsistent author name variations), and the added complexity of working across multiple languages pose further challenges.
The Digital Collections Manager
Goals: Sergio wants to avoid data silos and increasingly aims to use AI as a tool for exploring data in innovative ways.
Needs: To work effectively, Sergio needs access to technical training, best practice guidelines and documentation to build his confidence. He needs help with tools and workflows for entity recognition (NER or MNER) for images and image captions, item descriptions, and abstracts.
Pain Points: Sergio regularly faces issues with poorly structured, incomplete or multilingual data. Transforming data for research needs is complicated by the multitude of data formats. API access limitations or lack of information on available tools pose further challenges.
The Digital Humanities Researcher
Goals: Astrid’s primary goal is to reduce the manual effort in her day-to-day tasks. She wants to ensure that local entities are being matched accurately to the right Wikidata IDs. Access to data usage analytics is a further goal.
Needs: Astrid needs optimised workflows for image and data submission via API, and build-in deduplication capabilities. She’s looking to AI to help generate additional metadata — such as inferring gender from first names — but only where it can do so reliably and without hallucinations.
Pain Points: Astrid frequently has to deal with unstructured or messy data, duplicate IDs and incomplete or missing rights information. The preparation of Wikitext for Wikimedia Commons is another time-consuming task. Most challenging is creating new Wikidata entities from scratch when there’s little to no supporting data to work from.
The Wikimedian-in-Residence
The Museum Curator
Goals: Nina wants to increase data findability and better support cross-departmental work in her institution. Enhancing existing collection data with external resources is also important for presentation purposes (e.g. in museum displays, not just back-of-house). She aims to follow up data exports to Wikidata with editorial work on Wikipedia.
Needs: Nina needs guidance on ontology modeling and practical advice on data enrichment. She would like to work with semi-automated data entry workflows and use standard schemas. She also needs statistics on how collection data is being accessed and used.
Pain points: Nina is held back by the manual effort required by current data curation processes and the lack of automation. Further challenges are posed by limited staff and resources, skills gaps, and inflexible legacy systems.
The Digital Collections Manager
Goals: Andreas wants to share collection data seamlessly across multiple platforms and data aggregators, and link to authority files where relevant. A core ambition is reliable roundtripping – being able to publish data and pull back crowdsourced erichments without losing integrity.
Needs: Andreas needs structured workflows for data deduplication and transformation, terminology mapping and entity linking. Data validation is also a requirement. Automating the alignment of local entities to Wikidata IDs is a particular priority for the roundtripping process.
Pain points: Andreas struggles with the limitations of tools like OpenRefine, which don’t scale well for his workflows. The absence of real-time data sync with some platforms adds friction, while data silos make it hard to get a unified view. Incomplete records, messy data (e.g. inconsistent author name variations), and the added complexity of working across multiple languages pose further challenges.
The Digital Humanities Researcher
Goals: Sergio wants to avoid data silos and increasingly aims to use AI as a tool for exploring data in innovative ways.
Needs: To work effectively, Sergio needs access to technical training, best practice guidelines and documentation to build his confidence. He needs help with tools and workflows for entity recognition (NER or MNER) for images and image captions, item descriptions, and abstracts.
Pain Points: Sergio regularly faces issues with poorly structured, incomplete or multilingual data. Transforming data for research needs is complicated by the multitude of data formats. API access limitations or lack of information on available tools pose further challenges.
The Wikimedian-in-Residence
Goals: Astrid’s primary goal is to reduce the manual effort in her day-to-day tasks. She wants to ensure that local entities are being matched accurately to the right Wikidata IDs. Access to data usage analytics is a further goal.
Needs: Astrid needs optimised workflows for image and data submission via API, and build-in deduplication capabilities. She’s looking to AI to help generate additional metadata — such as inferring gender from first names — but only where it can do so reliably and without hallucinations.
Pain Points: Astrid frequently has to deal with unstructured or messy data, duplicate IDs and incomplete or missing rights information. The preparation of Wikitext for Wikimedia Commons is another time-consuming task. Most challenging is creating new Wikidata entities from scratch when there’s little to no supporting data to work from.
Interested in taking part?
ECHOLOT is committed to continuous engagement with our stakeholder communities. We value regular feedback and run surveys, organise interactive co-design workshops and training sessions. If you’re interested in taking part in some of these activities, want to stay informed for future events, or just want to share direct feedback, do not hesitate to get in touch below! If you’re more hands-on and want to contribute bug reports or issues connected to our tool development, you can also directly visit our GitHub repositories and submit an issue there.