---

Introduction: A Real Challenge for the Team

Recently, a debate erupted in our Slack channel about how to simplify the onboarding process for new employers. One of our developers pointed out that many of them struggle with filling out the required information, complicating their integration into the system. The discussion lasted several days, and it became clear that this was an issue that needed serious attention.

Why This Matters

The onboarding issue for employers directly affected our matching team. We understood that a complicated process could lead to high dropout rates and the loss of potential partners. The job market is becoming increasingly competitive, and every day that a new employer cannot start working with us is a lost opportunity for both them and us. We decided that we needed to create a more intuitive and faster process.

The Problem in Detail

The existing onboarding process involved multiple steps that required employers to fill out lengthy forms with information about their company, job vacancies, and requirements. For example, one employer we spoke with spent over 30 minutes filling out a form, after which they had questions about how to correctly complete certain fields. This created a negative impression and reduced their interest in collaboration.

Initial Attempts at a Solution

At the first stage, we tried to improve the interface by adding hints and examples for filling out forms. However, this did not yield significant improvements. We noticed that even with hints, employers still faced difficulties, especially with complex wording and requirements. Ultimately, we realized that improving the interface was only a temporary fix, and we needed a more radical change.

Technical Approach

After several discussions, we decided to use machine learning to create an AI-supported onboarding process. We developed a model that analyzes the input data and provides real-time suggestions. Here’s a brief code example demonstrating how we process the input data:

class OnboardingAI:
    def suggest_fields(self, input_data):
        # Logic for data analysis
        suggestions = self.model.predict(input_data)
        return suggestions

This model allows employers to receive form-filling suggestions based on their previous actions and company information. We also integrated a chatbot that can answer questions during the form-filling process.

Product Changes

After implementing the AI-supported process, we noticed a significant reduction in the time required to complete onboarding. Employers could now finish the process in 10-15 minutes, improving the overall experience of working with our platform. We also received positive feedback on the /for-companies page, where employers share their experiences with Fitlane AI.

What We Learned

• Interface Doesn’t Always Solve the Problem: Improving the interface can help, but it doesn’t always address the root issues.
• AI Can Simplify Complex Processes: Utilizing machine learning for data analysis opened new opportunities for streamlining interactions.
• User Experience is Paramount: Quick and intuitive onboarding processes can significantly increase employer engagement.

What This Means for Candidates

For candidates, this means that the application process for job vacancies has become more transparent and faster. With the help of AI, employers can find suitable candidates more quickly, increasing the chances of landing the desired job. Candidates can now see how employers use the platform, thanks to improved integration and feedback.

What This Means for Employers

For employers, this means they can focus on more important aspects—finding talent and developing their teams—rather than filling out forms. The simplified onboarding process allows them to start working with Fitlane AI more quickly and achieve results.

Next Steps

Although we have already made significant progress, there is still much work ahead. We continue to monitor how our model performs and gather feedback from users. In the future, we plan to enhance the AI algorithms to make the process even more adaptive. If we had to start from scratch, we would have tested the initial versions of the system more thoroughly to avoid some of the issues we encountered at the beginning. ---