5 Reasons Why Teneo Studio And Agile Development Are Made For Each Other
One of the most popular approaches to creating software is agile development. In this blog post, we’ll look at why Teneo Studio, Artificial Solutions’ integrated development environment for creating human-like, highly conversational digital assistants and chat bots, is such a great match with this approach.
What distinguishes agile development is a focus on short chunks of work that deliver limited but fully working slices of code, as well as continuous testing along the development process. Consequently, requirements and solutions can evolve together in a way that is responsive to changing circumstances.
However, an agile methodology also has its pitfalls. Examples include the importance of having automated tests, and the workload of continuous refactoring. This is where Teneo Studio can offer invaluable support to the team, so without further ado, here are my top five reasons why Teneo Studio and agile are a match made in heaven.
- Modules and rapid development
In an agile project, refactoring is not so much a demand as a basic survival skill. As the project requirements evolve, the solutions must evolve with them, and that typically means modifying the code. Fortunately, Teneo Studio lets you build small, reusable modules, or “flows”, very rapidly via its graphical user interface. As requirements evolve, you can isolate these modules, change them, expand them, disable them or delete them.
- Deliverables with added extras
Working on sprints that each deliver a limited slice of functionality might sound great in theory, but how do you carry it out in practice when working with AI? With limited initial functionality, your digital assistant or chat bot might look more like artificial stupidity than artificial intelligence. Teneo Studio solves this problem by making it very easy to quickly program the chat bot to give preliminary, generic answers to areas that have not yet been fully implemented. Later on, this work can be expanded or simply removed, thanks to Teneo Studio’s modular approach. Also, Teneo comes with a large amount of built-in conversational knowledge, which means that the bot can understand general inputs from the first sprint.
- Automated tests
In an agile methodology, where functionality and requirements evolve with the project, it is essential to test how any changes affect knowledge that has already been built. Teneo Studio lets the team add tests, in the form of inputs that should be understood, to different parts of the chat bot’s knowledge. This helps the team detect any unforeseen consequences of their changes, and quickly fix them.
- The branch not taken
In her book “Agile!” Sasha Mobley writes that agile is “scientific at heart, because everything is approached as an experiment… If something isn’t working, you cut it out as soon as possible and move on.” Teneo Studio is ideally suited for this philosophy since it allows rapid development, quick deployment, and support for experiments. I already mentioned how Teneo Studio allows rapid development. Updating the knowledge of an existing chat bot is also quick thanks to a simple and intuitive deployment procedure. Since Teneo modules are structured as flow charts, with nodes and branches, it is very easy to conduct experiments to determine which bot behavior gives better results. We can test by randomly choosing between two branches with different behaviors, and afterwards measuring if there is a difference in the outcome you are interested in. Once the results are in, the chat bot’s knowledge can be quickly modified to deliver only the behavior with the best result.
- Collecting data and information
And let’s not forget the importance of data for monitoring results. The agile philosophy of requirement/solution co-evolution requires clear and measurable goals. Teneo Studio is part of the larger Teneo Platform with its dedicated Teneo Analytics suite. With Teneo Analytics free-form, contextualized conversational data can be sliced and diced, measured and monitored. Crucially, Teneo Analytics is not limited to pre-defined KPIs, but lets the team flexibly measure what is relevant to the project.