Growth creates challenges
Datamart is growing: more and more full-time and part-time employees are joining our team. This mix creates a fun dynamic in the team, but it also brings challenges. This is because team members work on different days, which makes it difficult to keep everyone informed about ongoing projects and tasks. This can sometimes lead to not using each other's expertise, simply because a colleague does not know who is doing what.
Furthermore, Datamart's growth requires professionalization: having a structure in organizing tasks and processes is our top priority. We chose to use Jira as a central tool. At Jira, we keep track of tasks and plan sprints so that we work efficiently on projects together and can achieve our goals.
Our growth raised a number of questions:
- How do we ensure that everyone is aware of each other and can help where necessary?
- How do we get everyone to embrace Jira?
- How do we stimulate the connection within the team?
AI as a connector
As a solution to these challenges, we developed the “Company Courier”, a tool that sends a weekly, AI-generated newsletter based on the tasks and projects completed during the week.
The Company Courier is built entirely in Python and uses multiple APIs to generate and send the newsletter. Via an API call to Jira, all this week's tasks are retrieved and stored in a dataset. This dataset, in turn, is the input for the Large Language Model (LLM) that is used to generate the newsletter. The LLM used is Anthropic's Claude 3.5. Via the Anthropic API, the dataset, together with a carefully written prompt, is used as input to the model. The output is a complete newsletter. This is automatically sent every Friday afternoon using the Sendgrid API via a self-made dynamic template.
Importance of prompt engineering
The success of an AI solution such as the Company Courier depends on a well-thought-out prompt. Without clearly formulated instructions, you cannot trust that the quality of the model's output is always good. The prompt not only had to include the right structure for the newsletter, but also provide just enough context so that the newsletter shares relevant and interesting insights with the team. Before you can really use AI for these kinds of tasks, it's important to keep testing and trying. Be strict about assessing the output and keep going until you reach a prompt that ensures that the output is consistently usable and of a high standard.
The result
Every week, everyone receives, all employees receive an overview of what happened during the week, their colleagues did. This not only provides updates within the team, but also inspiration. By seeing what others are working on, new ideas and collaborations can arise, for example.
In addition, the Company Courier motivates team members to actively use Jira: only the tasks that are kept in Jira are included in the newsletter. This encourages everyone to keep track of their to-do's. You also want to know what the other person is up to, right?
Every week, everyone receives, all employees receive an overview of what happened during the week, their colleagues did. The Company Courier ensures that:
✅ Everyone in the team uses Jira consistently
📰 The team is always up to date with what everyone is doing
💡 Collaboration is stimulated and colleagues are inspired by the work of others
The newsletter that the Company Courier generates ensures that Datamart's progress and projects are just as on top of mind among all employees just before the weekend. This is how they start their well-deserved weekend inspired and fully up to date!
Curious about how AI can support your business?
Does your team also run into this problem? Get in touch to get the tool gratuitous to try!
Do you have another problem where we can help with the help of a smart AI tool? Then send a message now!
Altijd op de hoogte blijven? Laat hier je e-mail achter!
Blijf altijd op de hoogte van de laatste ontwikkelingen omtrent de mogelijkheden van data en 🤖 AI!