Case StudyHow Hâpy Services transformed its sales team through artificial intelligenceThe story of Clément GUTIERRES, the CEO who freed his team from the sales bottleneck with Antoine, His AI assistantDiscover the inspiring testimonial of Clément Gutierrez, CEO of Hâpy Services, who scaled his 10-person team into an augmented force of 22 collaborators (10 humans + 12 AI agents)!The challengeAn overwhelmed CEO, a stalled teamClément GUTIERRES leads Hâpy Services, a company specializing in connecting and securing remote equipment for its clients, using cellular and satellite networks. Based in La Roche-sur-Yon with a 10-person team, the company operates in a highly technical sector where every product reference, every price, and every specification matters.After 20 years at a Vendée-based industrial IT company, where he opened branches in Spain and Italy, Clément decided four years ago to take service to the next level. He co-founded Hâpy Services to provide clients with an additional layer of support in communication solutions.But very quickly, Clément faced a problem that threatened the growth of his company: he had become the bottleneck for his sales team."I wasn’t convinced by artificial intelligence. I was actively looking for a solution to a very real problem."Clément GUTIERRESCEO, Hâpy ServicesThe ProblemAll the essential sales information was centralized with him: international SIM card price lists, country-specific rates, technical specifications of connectivity solutions… His sales team depended entirely on him to respond to clients. When Clément was in a meeting, traveling to a trade show like SIDO in Lyon, or simply busy with a strategic project, the entire sales team slowed down."Our sales team needed to be augmented. The questions directed to me depended on my availability. It created a bottleneck" says Clément.This situation was unsustainable. Every unanswered question meant a potentially frustrated client. Every delay in response was a sales opportunity at risk of slipping away. For a growing company working with remote equipment worldwide, this structural bottleneck quickly became critical.The methodAn innovative approach: Start with the tasks, not the technologyFaced with this challenge, Clément took an unconventional approach. Unlike many entrepreneurs fascinated by the AI wave, he wasn’t looking to “do AI” just for the sake of it.“It’s a wave we all saw coming. We asked ourselves, as a team, how we could use it to provide better service to our clients,” he explains. “By investigating a little, we realized there was an obvious application: reducing our repetitive tasks and limiting tasks that barely leveraged our expertise.”Internal reflection without talking about AIClément initiated an internal reflection without even mentioning AI at first. He asked each team member, in their respective roles, to identify the tasks they would like to stop doing, or at least those that took up too much time and they wished to reduce.Identifying time-consuming tasks“Each person mentioned to us their top three tasks that took the most time and relied least on their expertise. Based on these insights, we created the priority list of agents to develop.”A pragmatic and participatory approachInstead of imposing a top-down technology, the team identified their own pain points. Team members didn’t experience change passively; they helped define what would actually assist them.“At the beginning, we didn’t talk about AI—we talked about use cases. Team members shared what was bothering them. The idea was to tell them they would no longer have to do those tasks. Naturally, there was acceptance, because it was beneficial for each team member.”Clément GUTIERRESCEO, Hâpy ServicesThe solutionMeeting Swiftask: The right tool at the right timeTo bring this vision to life, Clément tested several artificial intelligence solutions: ChatGPT, Claude, and even a proof of concept using an open-source model with Llama internally.But it was Swiftask that made the difference. “When I discovered Swiftask, I thought, let’s go for it ” recalls Clément.Why this choice?Access to all AI models“At Swiftask, we have access to all public models via APIs. Depending on the agents and the benefits we’re looking for, we can choose Claude, ChatGPT, or various other models. There are plenty of models available.”Data sovereignty and security“We wanted to keep part of our data private and not necessarily feed it into the public models of major providers like ChatGPT or Anthropic. What was interesting was the ability to choose semi-sovereign models.”Ease of use without coding“It’s truly business-oriented. You can even do it by voice. You can dictate. You don’t need to code.”A coherent business model“All of this for just 15 euros per subscription, in one place, with the same logic. It was primarily the ease of use, along with the option to have private models, that made the difference.”The AI agentAntoine, the new virtual colleagueWith Swiftask selected, Clément and his team created their first agent: Antoine, the sales expert.“Today, at Hâpy Services, we are ten humans and are assisted by twelve agents,” Clément proudly announces.Who Is Antoine?Antoine is a sales expert. His role is to provide support to the sales teams at Hâpy Services, explains Clément.Unlike a customer-facing chatbot that interacts directly with prospects, Antoine is an internal assistant. “My goal is that our clients don’t deal with AI. Everyone should make decisions based on their role. In our case, we decided that our added value is ourselves. Clients only deal with us. The idea, however, was to augment each team member.”How was Antoine created?The method is remarkably simple and effective.Essentially, this agent and its databases were built from the questions the sales team used to ask me. We loaded it with all of the company’s sales knowledge: commercial presentations, semi-technical presentations, and price lists.But the real work began afterward: training.Once everything was loaded, we asked Antoine lots of questions. Then we trained it to answer better and better. I personally handled this. Every question the team asked me over the course of two months, I also asked the agent. I observed how it responded, and if it didn’t answer like I would, I corrected it: “No, you’re wrong here.”Antoine’s brainIn Swiftask, each agent has a “brain”: its instructions, skills, and data.InstructionsThey define his role: we tell him who he is. Exactly as if you were telling someone, “your role here is to do this and that.” We give him his role and the expected functions.SkillsThey define what Antoine can do:✓ Search the internet✓ Create Excel files✓ Remember all feedback to improve future answersDataThey form his knowledge base: website content, webinar transcripts, price lists, SMS pricing, decision-support matrices based on solutions, client scenarios, configuration guides, the sales deck…“You can upload a PDF directly from your computer, or pull it from Drive, or even fetch it via API from external databases.”The resultsConcrete and measurable resultsSignificantly increased responsiveness“The impact is much greater responsiveness. Before, these were questions directed to me and depended on my availability. Now, the bottleneck is gone—or at least significantly reduced.”Greater team autonomy“For most questions, they now have the answers right there. They only come to me for very specific cases. For them, it’s an advantage because it avoids waiting. They’re more responsive and more autonomous.”Continuous learning“They also learn through Antoine, in a way, because a question is usually asked only once. So over time, they use it less and less.”A revolutionary onboarding“As soon as there’s a new team member, they have support immediately and become autonomous very quickly. Even though there’s a full training process, from day one, the person can handle an email by simply asking a question.”Freeing up the CEO’s time“It has freed up my time. I can focus on clients, strategic development, and partnerships, instead of spending my days answering questions about pricing.”An augmented team without hiring“To augment a sales team means increasing capacity with an additional agent. It’s like having one more salesperson alongside the agent, without going through a hiring process.”A staggering speed of creation“Creating an agent, if it’s not complex in terms of functionality, takes 5 to 10 minutes. You can experiment very, very quickly, which also allows room for trial and error.”"It’s truly a game changer."Clément GUTIERRESCEO, Hâpy ServicesThe RecommendationsAdvice from a pioneerBuilding on this successful experience, Clément shares his recommendations for other leaders considering this journey.1 Don’t talk about AI“The advice I would give for getting started with AI is not to talk about AI. Really focus on repetitive tasks, time-consuming tasks, and low-value tasks. And first, make a list of them.”2 Create a first agent quickly“From that observation, create a first agent, then a second, using Swiftask, and see how simple it is—once you know what you want to implement.”3 Involve the teams from the startBy asking each team member to list their most tedious tasks, Clément transformed a technology project into a human-centered project. Team members don’t resist the change : they become its drivers.4 Prioritize data sovereignty“We wanted to keep part of our data private. Data sovereignty and security are important for companies.”5 Choose a tool that allows experimentation“It allows room for trial and error. You can experiment very, very quickly. The ease of creating and modifying agents lets you learn by doing, without major risk.”The visionA clear vision of aI in businessWhat makes Hâpy Services’ story particularly inspiring is Clément’s clear and balanced vision of artificial intelligence.“I have one goal: our clients don’t deal with AI,” he affirms. “Everyone should make decisions based on their role. In our case, we decided that our added value is ourselves.”This stance is remarkable. While many companies rush to deploy customer-facing chatbots, Clément chose the opposite: AI augments his internal teams, but client interactions remain human.“Clients only deal with us. The idea, however, was to augment each team member.”This philosophy of AI as a tool for augmentation rather than replacement is the key to success. Team members don’t see Antoine as a threat—they see him as a colleague who helps them perform better.It’s not that a company can’t operate without AI, but thanks to AI, certain tasks can be done much faster.Clément GUTIERRESCEO, Hâpy ServicesConclusionKey TakeawaysThe story of Hâpy Services shows that adopting artificial intelligence in a company is neither mysterious nor complicated. It’s not about revolutionizing all processes overnight or becoming a “100% AI” company.It’s about:Identify repetitive tasksThat Consume Time and Require Little ExpertiseInvolve the teamsIn identifying needsChoose a suitable toolThat respects data sovereigntyCreate quicklyA First agent to experiment withIterate and improveBased on user feedbackMultiply the use casesOnce the first success is confirmed"We didn’t talk about AI, we talked about use cases."Clément GUTIERRESThis phrase could be the motto for any company aiming to successfully navigate digital transformation.The sales team is more autonomous, efficient, and responsive. New hires are operational from day one. Clément himself can focus on strategic business development instead of answering questions about pricing.And all of this is achieved without hiring a single additional person, with an investment of just 15 euros per user per month, while maintaining full control over sensitive sales data.Antoine, the virtual sales expert, doesn’t replace anyone. He enhances everyone’s capabilities. Perhaps this is the true promise of artificial intelligence in business: not to replace humans, but to free them from repetitive tasks so they can focus on what truly matters—client relationships, professional expertise, and creativity.Ready to transform your team with AI?Join Hâpy Services and hundreds of companies that have already boosted their capabilities with SwiftaskTalk to an AI expertFree trial