How Floom is Redefining the Rules of ML-based Innovation
Are you tasked with navigating the intricate labyrinth of developing machine learning-based applications? As designers, developers, data scientists or innovation managers, being assigned to coordinating multidisciplinary efforts can be laborious, ambiguous, repetitive and less than fun. The solution for all that? Floom — a groundbreaking visual innovation framework designed for user-centered machine learning within interdisciplinary teams. Created in collaboration with MHP — A Porsche Company, Floom enhances communication and boosts efficiency in innovation processes for machine learning applications.
Floom’s unique capabilities stretch across a spectrum of projects. From customer consulting to streamlining internal innovation, it all starts with a spark of an idea. This cutting-edge framework navigates agile teams through every step to transform that idea into a tangible intelligent MVP.
The standout feature of Floom is its visual interface. By representing each team member as a color-coded avatar, you can track individual progress through a sequence of rooms. Each room hosts a set of tasks, bolstered by resources and solution guidelines, making the innovation journey both accessible and engaging.
Fostering collaborative brainstorming, Floom earmarks group work spaces within its layout. Here, the collective efforts of each member are assembled on interactive whiteboards, forming a unified “Source of Truth.” This innovative approach eliminates knowledge silos and encourages a cohesive progression towards a shared goal.
Floom is a great innovation management guide. It promotes transparency, fuels internal communication, facilitates knowledge sharing, and cultivates a productive team dynamic. Floom also empowers managers and owners to prioritize effectively, manage the project backlog, and maintain the product vision. In addition, it enables the smooth delivery of project results to stakeholders, ensuring that the team’s hard work gets the recognition it deserves.
Floom’s standout feature, however, is the introduction of worksheets, bridging the gap between designers, data scientists, and developers. These worksheets interpret user needs into data requirements, helping to define appropriate models and sculpt an intelligent user experience. This guarantees each discipline’s unique perspectives are incorporated into the product development process, enhancing overall effectiveness.
Floom is built with LabKit, a massive system of components to build virtual spaces for collaborative experiences in Figma. Due to the modal design approach, organizations and teams can change the current innovation process Floom provides and easily tailor it to their needs.
Now, how does Floom work in practice? Imagine your team is tasked with an exciting challenge: developing a Minimum Viable Product (MVP) for an airline with the goal of reducing in-flight meal waste. It’s an ambitious task that requires an interdisciplinary approach. This is where Floom shines.
At Floom, every project starts with the business hypothesis, the starting point for every venture. In this case, the hypothesis might be: “We believe that by predicting passengers’ meal choices and preferences using machine learning, we can significantly reduce in-flight food waste, improve customer satisfaction, and ultimately save money for the airline.”
This hypothesis sets the stage for the work to come. It’s a shared starting point that guides your team, unites them around a common goal, and allows everyone to see the big picture.
Next, we move to the brainstorming stage. Here, all team members, from designers and developers to data scientists and business experts, come together to generate ideas on how to achieve this goal. Using Floom’s interactive framework, team members are represented as avatars moving through various rooms, each with distinct tasks and objectives. This keeps everyone in sync and aware of where they stand in the process.
With a list of ideas, we’re ready for the next phase, where designers and data scientists work together to translate these concepts into tangible user and data needs. This collaboration allows the team to define the requirements of the machine learning model, ensuring it’s tailored to the real-world needs of the users.
Once the user and data needs are defined, the team moves into the prototyping phase. This is where the MVP begins to take shape. The team designs and develops a fast prototype, enabling the entire team to envision the product and provide immediate feedback.
Finally, the team takes the fast prototype and works together to build, test, and refine the MVP. This is an iterative process, where feedback is taken on board, modifications are made, and the product is continuously improved. Once the MVP is ready, it’s released to the market for the client, ready to change the way airlines manage in-flight meals.
Throughout this entire process, Floom ensures each team member knows their role and responsibilities, facilitating a transparent and efficient workflow. It creates an environment where every voice is heard, and every discipline is valued.
In conclusion, Floom is a customizable end-to-end framework designed to optimize the development of user-centric machine learning applications in interdisciplinary teams. By providing a comprehensive, easy-to-navigate platform, Floom enables collaboration and productivity in innovation processes.
FLOOM in a nutshell
- Floom is the first visual interactive innovation framework for human-centered machine learning
- Floom is an end-to-end solution for agile innovation teams
- It is open source and free of charge
- Floom is fully customizable to any organization’s and team’s needs
- It ensures project progress visibility and shareability of all tasks and results
- It accelerates the complex machine learning innovation process and makes it more transparent
- Floom integrates agile approaches like Scrum and Design Thinking into a single framework
- It facilitates interdisciplinary project work and promotes enjoyment and team spirit
