How I Took a 4-Month Maternity Leave with No Backfill, Thanks to AI

In the fast-paced world of fintech, taking maternity leave as a key player in a small startup can seem daunting, especially when you’re the one person responsible for the entire customer experience—from onboarding to offboarding. With a growing customer base of 10,000 small businesses and over half a billion euros in loans financed, the challenge wasn’t just to maintain operations but to improve them during my absence. 

How did I manage to take four months of maternity leave with no backfill? The answer lies in the strategic deployment of AI, in addition to the wonderful support of my colleagues!

The birth of our AI team

Faced with the need to cover multiple roles, we introduced three AI assistants—Emma, Amelie, and Alexis—to ensure that our customer experience didn’t just survive but thrive.

• Emma was tasked with educating prospects and guiding them through the sign-up process, making sure potential customers understood our offerings and felt confident in moving forward. For example, Emma handles questions about our pricing and loan terms, and reassures our customers about our security and compliance requirements as a regulated financial institution. 
• Amelie took charge of customer onboarding and activation, focusing on ensuring that new customers were not just onboarded but also actively using their loan facility, which is crucial for our revenue stream. This includes educating customers about the types of invoices they can submit for financing, and escalating bugs or errors to human support.
• Alexis handled internal operations, streamlining processes for the team by surfacing the correct procedures and maintaining our brand’s tone and voice in all communications. One of the biggest impacts Alexis has had is drastically improving the efficiency of the team in finding the appropriate resolution process, instead of searching a knowledge base with hundreds of articles. Added bonus: the quality of our support responses is more consistent. 

Onboarding our AI assistants in less than 2 weeks 

Rolling out our AI assistants—Emma, Amelie, and Alexis—was a testament to speed and precision, achieved through thorough testing and constant iteration. We knew that quick deployment wasn’t enough; these AI tools needed to be fully equipped to handle the complexities of our customer interactions and internal processes.

1. Outlining a thorough testing plan:

We knew from the start that if we wanted this to work, we had to nail the testing phase. So, we put together a comprehensive plan that tackled all the possible scenarios we could think of.

• Simulating Real-World Scenarios: We didn’t just play around with the AI; we threw real customer interactions at it, from easy questions to the really tricky stuff, to make sure it could handle whatever came its way. 
• Stress Testing: We cranked up the volume of queries to see if the AI could keep up under pressure without breaking a sweat. It needed to stay sharp and accurate, even during peak times.
• Feedback Loops: We set up a continuous feedback loop, letting early users give us their honest opinions on how the AI was doing, as well as scrunitising each conversation that fell into the inbox. We used that input to keep tweaking and improving the system as we went along.

‍

2. Dogfooding AI at every step

To guarantee that our AI was up to the task, we embraced the concept of "dogfooding"—using the AI ourselves before rolling it out to our customers. By being our own first users, we could fine-tune the AI in real time, ensuring it was not only functional but also aligned with our brand’s tone and voice. This meant we used ChatGPT throughout the entire process, from analysing over 2,000 customer conversations to identifying key themes, suggesting responses, refining our tone, and recommending additional content.

Transformative AI for better CX, business, and peace of mind

The impact of AI on our workflow was transformative. We slashed customer resolution time from half a day to just 22 minutes, dramatically boosting our responsiveness. Loan activation rates jumped by 24%, which directly increased our revenue. And with AI running 24/7, we accelerated our learning curve, quickly processing and analysing large amounts of data to gain deeper insights and make more informed decisions.

Our AI team not only improved our operational efficiency while I was out, but upon return put me in a better position to understand our customer needs to enhance their overall experience. 

When I returned from leave, the time saved by AI allowed me to shift my focus to more strategic activities. With routine tasks automated, I could dive deeper into analysing customer behaviour, leading to better insights and strategies. This freed me up to lead larger projects and drive more impactful initiatives across the company. AI didn't just keep the wheels turning; it propelled us forward.

The vital role of human oversight

Human oversight is crucial in our AI journey, ensuring that the technology we rely on enhances, rather than replaces, the human touch that defines our customer relationships. During my mat-leave, we implemented a system where team members took turns reviewing Amelie’s work and asynchronously responding to customer needs. This wasn’t just a stopgap measure; it became a valuable practice that we’ve continued to this day with every Defacto teammate doing customer care at least once a week. 

‍

‍

By having different team members regularly check in on Amelie’s outputs, we maintain a strong connection with our customers, ensuring that their needs are met with the same empathy and understanding as before. This practice also accelerates learning across the team, as each member gains insights into our customer-facing product and customer feedback.

This blend of AI efficiency and human oversight creates a careful balance where AI serves as a powerful tool to streamline operations, while human expertise ensures that the personal touch and trust we’ve built with our customers remain intact. 

Looking forward: The future of AI and our commitment to trust

As we move forward, we're excited to explore new possibilities with AI, particularly in personalising our interactions and developing industry-specific content that truly resonates with our diverse customer base. AI has not only been a time-saver but has also empowered our team to work smarter and more efficiently.

On a personal level, I’ve embraced AI in nearly every aspect of my work, from querying SQL for customer insights to organising my thoughts for this very article. It’s become an indispensable tool that’s made my one-woman operations team a scalable possibility. 

However, innovation cannot come at the cost of trust. As we continue to leverage AI as a powerful tool for continuous improvement, we remain dedicated to upholding the highest standards of compliance and security. AI has proven to be a game-changer, not just during my maternity leave but as a critical component of our strategy to enhance customer experiences and drive growth.

As AI reshapes how we operate, we're committed to ensuring that it continues to support our mission of delivering a trusted and exceptional customer experience. 

If you’re exploring innovative ways to use AI in your business, we’d love to hear about them!