How Alexandre, Product Operations at PayFit, Uses Dust: A Sneak Peek at His Custom Assistants
Discover how PayFit leverages Dust AI assistants to improve their internal processes.
PayFit, a French unicorn specializing in payroll and HR management, is using AI to improve internal processes across different functions and domains of expertise. A member of the Next40, with teams in France, the UK, and Spain, PayFit supports over 16,000 businesses and manages the HR needs of more than 220,000 employees.
As the technology arose, the company recognized the potential of generative AI to enhance operations and empower and augment Payfit's teams. They worked on identifying relevant use cases and solutions and equipping everyone with the knowledge and tools to leverage AI.
By rolling out Dust and encouraging experimentation, PayFit aimed to facilitate the adoption of GenAI, enabling employees to augment their skills, focus on what matters, and contribute to innovation in their roles.
🌎 PayFit, launched in France and offering its services in France, Spain and the UK, is a leading SaaS company that provides payroll and HR management for thousands of entrepreneurs and small and medium sized companies. By automating complex processes, ensuring compliance with local regulations and a modern user experience, PayFit simplifies payroll and HR tasks, allowing businesses to focus more on their core activities.
How it all started
PayFit’s deeper exploration of generative AI really began in the summer of 2023. They created the first internal community around AI topics, with guidelines for people to test tools while respecting privacy and confidential information. Beyond playing with ChatGPT and a few other tools, they decided in the summer of 2023 to dive deeper into exploring the potential of AI focused on key use cases relevant to them, both internal and customer-facing. The core team started small with Ghislain, one of PayFit's co-founders, Alex from the product operations team, and Sylvain, the CPO.
They knew AI could be a game-changer and that, beyond the incredible improvements made by the Large Language Models (LLMs), the magic was in the use case application and integration into key routines and processes, with the potential to show the first improvements quickly.
With Dust, they aimed first to free up creative energy for more strategic initiatives by automating repetitive tasks and streamlining information retrieval.
Their goal was clear: speed up operations, starting with a pilot with a subset of the customer success team in France and unlocking new use cases across the company.
"It started with Ghislain, Sylvain, and me exploring how AI could help the business, we knew the importance of focusing on specific use cases, and the need to have a platform that supports collaboration and adoption" - Alex
Recognizing that different departments had different needs, Alex, Ghislain, and Sylvain brought in key team members, “AI relays or champions,” from various functions with their deep “expert understanding.”
They worked collaboratively to develop internal AI assistants tailored to meet diverse team requirements, including specific tools for the teams to increase their productivity and effectiveness.
Content Creation with AI
Better Manage Meeting Recaps
One challenge the marketing team faced was keeping everyone updated on meetings, especially when team members couldn't attend.
They tackled this problem by creating a Dust meeting recap assistant. With this tool, the team could record meetings, upload the transcripts, and get a quick summary of all the important points discussed.
The assistant has made it much easier for the team to stay aligned and act on decisions quickly.
"What was a big surprise to me was that you can have a good assistant in less than 30 seconds." - Alex
Content Creation
Another significant improvement addressed the challenge of producing high-quality content for customers more efficiently.
They developed a Dust assistant that could draft the initial version of an article based on a prompt and internal product knowledge. Now, the marketing team can generate content drafts in just a few clicks and focus their efforts on refining and perfecting the final version, beyond focusing even more on the brief.
“Now we can use AI to generate the first draft of an article that will be then published to customers.” - Alex
Improving Job Postings
Alex’s work to improve marketing efforts extended to collaborating with the talent acquisition team, where he developed an AI assistant to improve job offers. By analyzing existing job offers and providing valuable feedback, the assistant helped refine the content and make it more appealing to potential employees.
The assistant’s insights impressed the senior talent acquisition staff, acknowledging that it highlighted important points. This assistant not only enhanced the quality of job offers but also demonstrated the potential of AI to bring a fresh perspective to marketing materials.
"The advice it gives and the points it notices are really valid and valuable. It is a great companion that helps reach a better outcome." - A PayFit senior talent acquisition officer.
Knowledge Base Search Assistant
They also addressed the challenge of quickly finding specific information within a vast internal knowledge base. They developed a knowledge-based search assistant that allows teams to ask questions and receive instant, accurate answers. This tool eliminates the need to manually sift through internal documents, which saves time and boosts efficiency.
By streamlining information retrieval, the assistant ensures that teams can focus on their core tasks without getting bogged down by lengthy searches.
"The importance of ease of use in AI platforms like Dust cannot be overstated. The real magic happens at the interface between people, expertise specificities, company data, and LLMs. By enabling seamless integration with internal knowledge bases, Dust helps bridge the gap between isolated information silos. Combining the robust capabilities of AI with the nuanced expertise inherent in specific domains unlocks unprecedented potential, driving efficiency and innovation." - Sylvain
Future Directions and Employee Empowerment
The Dust assistants that we highlighted above were just the beginning at PayFit.
The company is now focused on expanding AI education across all its employees. The goal is progressively to enable everyone at PayFit to confidently create their own Dust assistants, write prompts, and link their own data sources.
"Every position in the company will be impacted by AI and every position will benefit from it. So we better make sure they start using it as soon as possible.” - Alex
This initiative helps employees unlock numerous possibilities for automating tasks and integrating AI into their daily work. Ultimately, it empowers them to tailor Dust to their specific roles and needs, potentially transforming operations across the entire company.
Recommendations from Sylvain, Ghislain, and Alexandre:
📌 “Start experimenting with AI even if you don't feel fully ready. It's great at sorting information and gets better as you tweak it along the way.” - Ghislain
📌 “Take the time to really get to play with the technology. Understanding it well from the start and having a clear plan can make your AI tools much more effective.” - Alex
📌 “This represents an incredible learning opportunity. Play with it, think about how to leverage your data with the power of LLMs. Test different LLMs, update your prompt, work on your data sources, test, learn, repeat.” - Sylvain
PayFit's use of Dust Assistants highlights how AI can solve practical challenges across the business. Alex's experience shows AI's potential to streamline processes and foster a culture of continuous learning and adaptation.
This builder story also shows how companies like PayFit are leading by example, empowering employees to leverage technology and create more efficient and responsive operations.