More and more companies are organising Hackathons to accelerate innovation and create a corporate image that will attract developers. These coding competitions are designed to break down barriers to innovation within organisations, and promote a form of crowd-sourcing when open to the public.

At AT Internet, we planned our first Hackathon to coincide with our biennial internal event, the AT Summit. We decided to call it it “ShipIt Day,” since the idea was to deliver new innovations within 24 hours. In this article, we’ll share some insights into one of the projects we developed during this event. Along with five other colleagues, we formed a team tasked with working on the theme of “collaborative analytics tools”.

Collaborative methods are evolving

With the spread of data-driven decision-making and digital transformation experienced by many market players, the digital analyst is no longer isolated within his or her organisation, but has actually become the meeting point for the many needs within the company. We’ve recently made certain tools available to our customers to help them better share data throughout their organisation: the Reports and Dashboards applications. The democratisation of data is a need that’s fundamental but insufficiently addressed.

Usage of our solutions and working methods are evolving very rapidly within organisations and are becoming increasingly collaborative. The massive influx of millennials into the job market – these ultra-connected employees who prefer digital usage – is bringing radically different methods of interaction into our professional environment. The ecosystem of collaborative tools is exploding, and products that were unknown even 5 years ago are becoming the norm in terms of collaboration. For example, real-time messaging tools are taking precedence over email. Microsoft’s purchase of Yammer, and more recently, the nearly viral spread of Slack, are proof of the interest sparked by these new products.

Fertile ground for service integration

If tool usage is clearly changing, the application market is also undergoing a significant evolution that we might call the globalisation of services. We’re moving from an era where professions were siloed by software, to an era where myriad services are available via simple HTTP calls (such as your browser) – these are the famous APIs (application programming interfaces). This fundamental shift was started by Amazon with its manifesto for systems fully accessible by API. Today, the majority of tools and services available on the web offer integrations with other applications. The giants of the digital world are leading the way here: Twitter, LinkedIn, Facebook, and Google let everyone communicate with their products without having to go through the user interface. Even some smaller tools try to act as the glue between several SaaS applications to offer simple integration without needing to be a developer – for example, Zapier and IFTTT.

The latest wave of this ecosystem is built on the appearance of “chatbots”. These services have a little intelligence built in, and are capable of carrying out simple actions within the context of a discussion. Along these lines, Facebook recently announced the launch of its chatbot platform, which will be integrated with Messenger. The prospects for these chatbots are enormous; their most enthusiastic proponents see them becoming our personal assistants. Without going that far, we can easily consider these bots as a new approach to human-machine interaction in lieu of a standard graphical interface. We might just be experiencing the death of web design and its elaborate graphic components.

The ultra-collaborative Analytics Suite

We challenged ourselves to bring our Analytics Suite into the age of total digital collaboration while reusing – as much as possible – tools available on the market and web APIs.

ShipIt Day team hackathon AT Internet

In order to quickly achieve results and avoid ending up with nothing to show after 24 hours, we adopted an extremely iterative mode of operation:

  • Work sessions lasting 30 minutes maximum, during which each team member was free to work on whatever they liked, with whomever they wanted on the team, as long as they remained within the defined theme.
  • At the end of each session, a 5-minute briefing with the entire team. Everyone presented what they had done, their obstacles, and their future possibilities, and we collectively decided to abandon an idea if it didn’t produce concrete results after 2 work sessions.

ShipIt Day briefing development

Using this as a framework, we explored several different ideas, and these were the most promising – that is to say, those upon which we could base a convincing prototype:

  • An integration of our Analytics Suite with Zapier in order to trigger the storage of an export from our Dashboards application in Dropbox. We reused the notification system in the Analytics Suiteto trigger a Zap that fetches the link from your export and stores the document in your Dropbox space.
  • An integration between Slack and our Dashboardsapplication that lets you attach a discussion thread that comes from Slack to the menu bar of the dashboard. Each dashboard could thus have its own discussion thread in your Slack group and would let you work together either directly in the Slack interface or when you’re visiting your dashboard.
  • The creation of a chatbot that lets you perform simple commands in the Analytics Suite. We’ve given our chatbot the ability to carry out two operations. From Slack, you can ask our bot to export a dashboard by naming it, and you’ll receive a response from the bot with the download link for this document directly in your Slack discussion thread. You can also ask our bot to give you the number of visits for the previous day by telling it the name of the site, and it will respond with this metric in your discussion thread.

What are the possibilities?

The possibilities and opportunities offered by the integration of these different tools are very exciting, and we’ve seen how quickly it’s possible to provide collaborative services within our solution. The interconnections with Slack to communicate within the Analytics Suite would let us rely on turnkey tools to promote collaboration and avoid developing redundant features. One could imagine an interconnection with Yammer using the same approach.

The use of a chatbot could act as a real assistant to the team of analysts and – to a certain extent – replace the use of an interface.

Finally, the integration of different web services through tools like Zapier seems to be the most attainable in the short term by connecting to our notification system. We can imagine any type of integration with a web tool from events that occur in the notifications from our interface.

However, the most time-consuming aspect – and the one that will require a more significant effort – will be securing these services and enabling their operation on a large scale. It will also be necessary to pursue the technological openness of our platform to allow it to integrate more easily into the web ecosystem.

And you: What collaborative tools do you use in your daily routine at your company? Would you be interested in one of the projects outlined above focused on the theme of collaboration?


Gilles is Business Unit Manager within AT Internet’s R&D team. His teams are in charge of the Analytics Suite’s SaaS environment, the development of the Reports and Dashboards applications, and data bridges. An IT engineer, Gilles has more than 10 years of experience in the development of digital products and managing dev teams.

Comments are closed.