France

Innovators/Entrepreneurs

Date of the expedition

from 08/01/2024 to 07/07/2024

Selected Track

Open Ideas

Project title

Datami

Host Organization

Mitacs

Media

Biography

I am a software developer, and one of the co-founders of multi, a french digital cooperative.

I believe in decentralised organisation, collective intelligence, grassroot ideas, and in developing open source solutions for more resilient democracies.

After initial training as an architect, then having worked in international cultural cooperation (Mexico, Egypt), then in research field (a CNRS contract in Turkey), I finally converted to digital as fullstack developer.

Over the last 10 years I have mainly worked for ministerial institutions (Bercy, CGET, Agence Bio), inter-ministerial (DINUM), associations and think tanks active in the general interest (PiNG, Ternum, Rhinocc, Décider Ensemble), as well for local public structures. I participated in Etalab’s “General Interest Entrepreneurs” program in 2018, and continued as an independent, until I co-created the cooperative multi in 2022.

I participated in designing and developing multiple free digital tools (some having received awards or grants) for various purposes: data publication, data editing, data valorisation, and data visualisation.

Project Summary

Grassroot organisations still have difficulties sharing or maintaining their open data.

At a local level small or grassroot organisations have an important role in raising public awareness about topics of public interest (environment, social equality, urbanism, health…). Despite this, local institutions or small associations refrain from sharing the data they gather about a territory or a field of expertise due to a structural lack of means or technical skills. Open data produced locally are in that sense potential true decentralised digital commons. However existing tools are not designed for this kind of organisation, resulting in invisibilizing important knowledge of public interest.

Small organisations produce or may even publish raw datasets of public interest, but a gap remains between local data producers and their publics : a raw dataset may not be intelligible enough for a broad range of users ; open data may also need interoperability, curation or grassroot contributions to stay up to date so they could be useful for other application domains.

The project is called Datami, it is an open source solution to make grassroot open data more intelligible, more affordable, and to simplify civic contributions.

Datami is a system of modular web components of data visualisation, backed with external services for data management. Datami components can be freely embedded onto any website, while still benefiting from robust version control and moderation features.

Open data could be a real bridge between citizen and civic society. Datami aims to help a wide spectrum of organisations to produce and display their own data in the open, while staying fully in control.

Datami is also designed to simplify the act of contributing to open data. Datami aims to give broad audiences the taste to contribute to open data : no account to create, nice visualisations, simple interfaces, pre-structured data, a system of moderation… We noted that no other open source solution currently provides such moderation features for a broad audience as simply as we imagined.

As an open source project Datami needs use cases, relays and allies (individuals or institutions) in countries leading the way on open data policies such as Canada or France. Our strategy consisted in relying on new partnerships and use cases to improve our solution.

The NGI Enrichers program allowed me to stay in Canada for 6 months, in cooperation with Mitacs and Concordia University. The roadmap for this expedition was to assist a research project team from Concordia university and propose to them using or adapting Datami for the research project. The research project is coordinated by KEPO (Kahnawà:ke Environment Protection Office), and consists in monitoring air quality in the Kahnawà:ke indigenous territory (at the south of Montreal on the St Laurent river). The project title is “Air Quality Monitoring Network in Kahnawà:ke”, that we will refer to as AQMN in the next paragraphs.

The data collected in the AQMN project should be displayed online, in a way citizens (from Kahnawà:ke in particular) could intuitively understand the levels of pollutants in the air. The AQMN methodology to gather air quality datasets is also based on citizen participation.

The research team created air monitoring devices (Raspberry Pi) citizens from Kahnawà:ke could put at home or outside throughout the neighbourhood. These devices are connected to another Raspberry Pi server, on which all datasets are stored in base as time series. This same device is connected to the internet and provides an API to let external users or services fetch the raw datasets.

The AQMN project has the particularity to be designed as an “open science” project, meaning also data produced during the research project (as for digital developments such as custom scripts or applications) belongs from the start to KEPO.

The ambition carried by the AQMN project is to allow any citizen to access and intuitively understand the state of air quality, and therefore evaluate the incurred potential health risks.

We proposed Datami as a part of the technical solution to this challenge, as it respects the several constraints of such a specific and experimental project the AQMN project is. Datami ticks some critical “checkboxes” of AQMN constraints : open source licence, storage and control over the datasets by the data producer itself, data visualisation, portability onto external websites, and most simple and low cost solution for all stakeholders.

The project addresses the following topics : Decentralised Internet, Intuitive user experience, Data sovereignty, and Digital commons.

Key Result

The expected results we had in mind before the expedition were to search for and develop new partnerships between Multi and Canada with local entities, starting with Concordia and direct use of Datami, but also with other Canadian actors (public and private) directly interested in open data in a broad sense:  public sector open data, open science, data commons, etc…

Another family of expected results were the technical developments on Datami itself. Testing and speaking about Datami in a different cultural or commercial environment would lead to adapt or rethink Datami’s technical roadmap or features.

So we had listed several kind of outcomes and impacts in our application, back almost one year ago now :

  • Partnerships and field testing with local canadian open data producers ;
  • New core developments and iterations ;
  • Better visibility of Datami solution on an international level ;
  • Experimentations, improvements and feedback related to the solution’s business model.

At this date the results of the expedition could be also listed on the same order:

Partnerships and field testing with local canadian open data producers ;

  • The idea of implementing Datami for the AQMN research project convinced the members, including and more importantly, the indigenous organisation (KEPO) in lead.
  • The major factor and challenge while working on the AQMN project was the project governance itself. The initiative and the technical solutions researchers would present need to be firsthand considered and accepted by the lead organisation KEPO. Given the agenda or constraints of the consortium members, and also given some unexpected events (oil spill in the rivers of Kahnawà:ke for instance), organising video meetings and preparing the presentation took some time, before beginning any technical developments.
  • One of the main milestones on this topic was to prepare the presentation for KEPO with one of the researchers : to understand in depth the data infrastructure already in place, and also imagining several scenarios involving the use of Datami or not. This presentation presented 3 main scenarios and some variants, in order to propose an iterative roadmap and the certainty for KEPO to remain completely autonomous and sovereign over their datasets (even for a publication of it as open data).
  • The presentation we made to KEPO seemed to convince the partners to use Datami for data visualisation and its corresponding open source data management tools.
  • Still concerning the AQMN project, we are – at the time we are writing this report – just beginning to access the first exported datasets. The first developments (CI scripts, configuration, data processing…) are now barely beginning.
  • In addition to the partnership with KEPO we also approached several Canadian structures directly interested in open data policies / projects, namely : the Open North think tank, the library department of Concordia University, the library department of Montreal University, the think tank and fablab Sensorica, and the think tank Digital Transparency Lab. We presented a talk on Datami to an open science data work group in Concordia thanks to one of its librarians ; we presented another talk on Digital Commons business models in Sensorica, and we plan to present a talk in June with the Montreal University library’s open science work group.
  • We are currently discussing a grant application in partnership with Digital Transparency Lab, concerning a project involving Datami as a part of a digital infrastructure regarding Data Privacy.
  • We are also currently in the first steps of a commercial contract with Open North, regarding the Economic Valorisation of Data Commons.
  • As an unexpected outcome these advancements in Canada also permitted the multi cooperative to have more talks with French counterparts (Assemblée Virtuelle and the ActivityPods project). This led us to propose another application including Datami in the response for the NGI Commons program.
  • Last but not the least we – as the multi cooperative – applied to the NGI Search program 4th open call for the “DatamiPro” project (see below), a new grant for 123 000 € we obtained in April 2024.
  • To conclude this part we would add the several meetings we had in Montreal and the contacts made along the way raised up a central problem, common to every open source project indeed, but nonetheless acute for the Datami project : its business model. Datami being under open licence, led by a really small team, and still in its early developments, the efforts couldn’t always be made on the commercial front. There is a minimum threshold of users to reach so Datami’s economic model could be stable, but that needs upstream R&D. Datami’s R&D funding in the current economic environment is still possible mainly thanks to grants from the public sector, and will necessarily remain so in the next year at least. These thoughts explain why we decided in the first part of the expedition to be more proactive in searching for new partnerships and grant opportunities.
  • From the complicated economic situation projects such as Datami could have, we enlarged the scope of reflection to the more general question of “How to finance digital commons ?” and begun to add important chapters into a large presentation called “The Librery” documenting this reflection.

New core developments and iterations ;

  • As the first months in Montreal were mostly occupied by preparing and organising the meetings with KEPO on the AQMN project side and other Canadian actors of the open data scene, we also used this period to develop new strategies and test new technical hypotheses about the Datami project itself.
  • The several meetings we had during the first months showed there were some issues to address in priority in order to convince more entities to use Datami. From these meetings emerged the necessity to propose an even more “no-code” and “plug and play” web application than Datami currently is.
  • This confirmation in terms of needs and uses implied either A) to reshape the Datami project itself and its source code in depth, or B) to identify and design the missing interfaces and features, and then develop these apart from the original Datami project. We decided to test the later solution.
  • Thus, during this period we experimented with this approach and the intention to fit closer to Datami’s users needs : instead of planning technical evolutions in Datami’s core features, we started to develop a demo of an application “on top” of Datami we called DatamiPro. This independent application still uses Datami core data visualisation features, but addresses data “producers” more than data “consumers” : it aims to help data producers in preparing their datasets for open data, and help them upstream in configuring the Datami interfaces for publication.
  • The DatamiPro source code, even still experimental and not yet ready for production, is still published under open licence.
  • One of the main differences from Datami, the DatamiPro application includes an account management system, drag-and-drop interfaces to upload files, helpers to create interoperable data models, and Datami configuration and previews.
  • Another important difference from Datami is that we included into the DatamiPro system a way for users to subscribe (economically) to the service, the idea behind this being to experiment a more robust and recurrent business model than support currently is on Datami.
  • It is also notable that during the process of creating new contacts here in Canada and preparing project applications for NGI Search and NGI Zero Commons programs, we also made significant advancements regarding the technical roadmap of Datami.
  • In our NGI Search application we added a slide to resume the technical challenges and advantages of DatamiPro.
  • In our NGI Zero Commons application we developed another exciting idea : use ActivityPub protocol with ActivityPods P2P semantic system as a P2P independent backend for Datami and DatamiPro.

 Better visibility of Datami solution on an international level ;

  • Datami’s (including DatamiPro’s) expected business model is designed to be sustainable with a reasonable number of active users, a number that could be reached even within discrete communities such as the open data ones. By sticking to the same kind of community, but this time abroad, we aim to expand our main use cases and enhance Datami by this family of feedback.
  • We considered the experimental developments and meetings we mentioned above were full of insights. During the first two months we could advance on the technical side, while simultaneously we could have interviews with local open data actors and potential users of Datami. That alone led to making the product and its features more known to the Canadian Open Data or Open Science communities.
  • More concretely, these technical developments on DatamiPro and its roadmap were advanced enough for us to apply to the 4th open call of the NGI Search program. The ambition of DatamiPro is not less than being one of the funding pillars of the overall Datami project. Our application was fortunately accepted, and alone that would make our expedition in Canada a success.

Experimentations, improvements and feedback related to the solution’s business model.

  • Datami’s “legacy” business model was based on support for data preparation or for data Until now, we basically discussed using Datami face-to-face with our clients, client per client, project per project. Each project had its own peculiarities, and every mission included a few days for client’s data preparation, then a few days for UI configuration on Datami, and some days for coordination.
  • Whereas Datami’s source code remains open source, the cooperative multi was proposing a set of services around the solution : helping bootstrap the solution for a client, developing new features (participation to the roadmap)..
  • One of the most important feedback we had during the expedition (so far) about the preparation of datasets for open data by the actors we aim for, is the need to access a fully “ready to use and autonomous” service. The Canadian actors we met clearly emitted the need to use a version of Datami not relying on “support”, rather than directly use it as it is now – or fork it for an autonomous use.
  • The main improvement to the Datami business model was to give more emphasis to this need for fully autonomous uses, and therefore rethink the features set we could provide online with Datami in the next developments. This important evolution in the approach meant it was necessary to rethink the user experience, and led to the DatamiPro project.
  • Beyond the technical implications of this new strategy (developments and the cost of its R&D) DatamiPro makes a major leap in Datami business model, in the sense it makes all the project pivoting from support to SAAS, and for multi implies assuming a stronger position as software editor.
  • As we designed these major economical and technical evolutions taking care of remaining true to Datami roots : keep Datami under free open licence, while making its costs for users adapted to their means and needs. The subscription plans we will test for DatamiPro (as an online open data management tool) will address several types of users, and propose the most accessible prices for them  : free users with light needs ; around a dozen euros a month for small structures (NGOs, research labs, think tanks…) with more than 5 datasets to manage ; around fifty euros per month for bigger structures and more important needs…

Impact of the Fellowship

The expedition had in a relatively short period of time so far (approx. 3 ½ months) several impacts on the fellowship : in terms of technical developments, reorientation of the project’s business model, contacts and partnerships made along the way, fundings for the project, and an overall reflection on digital commons funding.

Technical developments :

Partnerships :

Funding Datami :

Funding digital commons :

This expedition has a deep effect on our way of thinking about digital projects in general, and will affect our idea of digital entrepreneurship itself. In only a couple of months  we noticed that – somehow more significantly than just sharing a language – we shared between France and Canada a strong common ideal for open data, extended to the role played by the civic society in the movement.

Open data

As a representative of multi we could talk with local actors about similarities and differences between our respective countries regarding open data, open and free source, digital projects life cycles, dependencies or not to government funding.

We first acquired the certainty that transatlantic cooperations could be possible on common projects between multi and Canadian partners, in fact some first cooperations are already in motion.

Business models

Our encounters with academics and open data activists here in Montreal also made us realise the Datami project was an interesting potential lever for traction and diversification for multi, a project that raised issues similar or deeper than in France (about open science for instance, or digital and territorial sovereignty).

In another hand, our discussions also led us to notice that important needs remained to be addressed, for Datami to become more immediately usable.

These reflections led us to the conclusion that another R&D cycle should open for Datami, that this cycle would require significant funding, and to formalise why and how this new cycle under the shape of the DatamiPro application.

 

Entrepreneurship

We also noticeably changed our vision of being a digital entrepreneur in the open source sector. So far, we considered our main value as entrepreneur was mainly our technical skills (such as coding, software architecture…). Nevertheless these skills during the expedition were comparatively less mobilised than organisational, conceptual or creative skills, in order to produce projects applications and presentations like we did since our arrival.

Future will tell but the work of fostering new partnerships and projects in Canada, added to the documentation of the core question of Funding Digital Commons, are comforting and engaging us in a direction of transforming our entrepreneurial practice.

 

Personal notes

On a more personal note we could add here that the upstream logistics and processes preparing the expedition were an adventure too (and still are). One of the conditions of being a fellow from the program was to remain an employee from our company. In our case it meant that at the same time we invested time and energy preparing the expedition for 6 months with my companion (logistics, household economics, and preparing Canadian partnerships upstream), keep working for previous company’s clients (including clients not directly involved in Datami), and advance on the expedition milestones (meetings and technical developments).

While respecting the NGI Enrichers program conditions, our responsibility was also to respect our company’s contract as an employee, in particular by working in parallel on another project for a French public administration (ADEME).  As a small company our cooperative can not afford to assume our full time salary without having a funded project (for the company) to work on, while the grant is concluded with the fellow as a physical person (i.e. not as an employee).

This “double job” situation during most of the expedition – to which we could add the jetlag with France – was tense and had sometimes led to moments needing adjustments. Even if internally at multi we discussed how to integrate the expedition to our work charge for 2024, the developments we made in Canada quickly made the work charge quite large.

Some adjustments were needed to be made in order to remedy some of these growing tensions with previous clients and/or our affiliated company multi.

In order to disengage progressively from the ADEME project to liberate more time on the NGI enrichers expedition, in march 2024 we had to reorient another developer (recently hired at multi) on the ADEME project as a replacement, thus also making necessary to implement a whole commercial and HR strategy to make that shift possible beforehand.

We simply wanted to point out this situation to strike out how to benefit from the NGI Enrichers grant (as a physical person) the necessity to remain an employee in a small company – in our case as a developer and partner in the cooperative multi – could lead to this challenging “double job” work load.

 

Conclusion on program impact

As we explained in detail above, the NGI Enrichers fellowship had important positive impacts on the Datami project : new use cases, new partnerships, new developments, new fundings.

But the expedition had also an impact on how our cooperative multi could address the  northern american market on topics we are specialised in as open data, open source, and digital commons. The consequences of this impact are still developing, but one of its first effects for multi is to assume more clearly the role of software editor in the near future.

In addition to the project and our company, the expedition will leave traces on our way to consider simultaneously digital entrepreneurship and activism in digital commons. The distance and differences from the Europe digital landscape made even clearer the necessity to push at the same time innovations that are technically acute, socially responsible, economically affordable and with limited profits, and rooted from the start in the civic society rather than government-led.

In a sentence we are reshaping our mission as an entrepreneur: helping to build a society of commons, beginning with digital commons.