The Humans Behind the AI: Leonard Fuechsel

1. Hi, Leo. Tell us about yourself! What was your path to joining Rasa?
Throughout my career, I’ve worked on building conversational AIs using various platforms. What I often found frustrating was how most of the big players treat natural language understanding like a black box - you can add training examples, but you can’t see or tweak what’s really going on.

That’s what drew me to Rasa first as a user: Rasa gave us the freedom to shape and refine the NLP pipeline ourselves. For our own custom smart speaker platform, we ended up choosing Rasa, specifically because of that flexibility. While working on this project also with non-technical stakeholders, I realized how much I missed having a proper UI while building our assistant - so when the opportunity came to join Rasa and help improve exactly that, it felt like the perfect next step for me.

2. Take us through a typical day on your role. What types of projects do you work on?

As a software engineer I like to start my day with some focus coding work.
In the morning we have async and sync standups and huddles with my team to share progress on current tasks, identify blockers and resolve them together. Weekly we sync with the rest of the tribe plan releases and reveal the roadmap.

Most of my day I spent on focused coding work, scratching and evaluating future features. Lately this was integrating Multi-Language Support in Rasa Studio. Currently we work on mocking custom actions and prompt editing in Rasa Studio. As all our engineers work on the entire stack, we can build whole features from scratch. Even so we work full remotely any peers are just a click away.

Regularly we share new learnings and ideas for improvements in dedicated meetings with brought groups. For my personal career we schedule 1:1 with my managers and close peers, in order to identify areas and next steps to grow.

3. What's an important problem you're solving at Rasa?
Enable the user to build powerful conversational AIs, which are free from hallucination and follow the business logic of the company.

4. How would you describe Rasa in three words?
Forward-thinking
Vibrant
Diverse

5. What’s the “wow” factor of Studio for you? And of your journey at Rasa?
What really excites me about Studio is how much easier it makes building conversational assistants. We’ve moved away from collecting rules, stories and training data. Now, we let the LLM handle conversations naturally, users can focus on describing their business logic in flows. Doing that in a visual interface is just so much more enjoyable than editing text files! Still, if you want to dive deeper or use more advanced tools, you can download your assistant and continue in your IDE.

6. What’s one thing you wish more users knew about Rasa’s capabilities?
Rephrase is such a powerful tool. It takes a predefined response, adjusts it to the context and previous conversation with that user by retaining the original message’s meaning.

7. What’s a feature or innovation you’re particularly excited about in Rasa’s future
Bring your own LLM on premise, we bring the finetuning recipes to work with calm
Provide voice channel out of the box

8. How do you collaborate with other teams at Rasa?
Currently I work in a team across products. While having regular syncs with the Studio team, also the Pro team members are just a slack message away. We regularly meet up to tackle new topics together.

9. What’s the most interesting thing you’ve learned lately?
Everyday I learn something new, what a LLM can do better. Using e.g., the rephrase with a multilingual assistant, makes your assistant respond in any language the user likes, without any efforts.

Thanks, Leo!

Want to team up with Rasa? We’re hiring! Find open positions on our Jobs Board.