Hey Makers
The wait is finally over, and we are excited to announce the winners of our recent community challenge! This time we have two absolutely fantastic, innovative, and truly deserving solutions that are making a difference. Meet @Soichiro_Nishihara and @Dupl3xx our winners and Masters of Make
Automated heating film order process
This solution was built by @Dupl3xx and his team.
What specific problem did you aim to solve?
We are a small company specialising in the sale of electric underfloor heating. Our customers send us floor plans of their houses, based on which our engineer prepares a complete design and quotation. However, this process was time consuming, customers had to wait a long time for a response and our designers were overloaded, often with unnecessary work.
Why was this particular solution necessary?
The project engineers spent a large amount of time preparing the bids, which resulted in several days of waiting for customers. This meant high costs for the company, which were often not recovered.
Who is the intended user of this solution?
The users are the customers themselves, who can generate the offer themselves via a form on our website, and also our designers, for whom this makes their work much easier.
Provide a brief overview of your solution and how it tackles the problem.
Our solution automates the entire process - the customer simply fills in an online form, selects the required heating film output and uploads a floor plan.
After submitting the form, the data is transferred via mailhook and then parsed. A three-step recognition process is then carried out using Anthropic Claude’s AI, which has proven to be the most reliable in testing:
Identification - AI recognizes the number of rooms, their types and total area.
Verification - The system checks the accuracy of the recognition and evaluates the accuracy rate in percentage.
Output generation - A complete summary of all key floor plan information is generated.
The resulting data is automatically sent to a pre-made template in Google Sheets that calculates the required material. The output is then exported to PDF, merged with other relevant documents, backed up to Google Drive and sent to the customer. The customer receives an immediate quotation and can order the material immediately via a link to the e-shop with a pre-populated shopping cart.
Bonus feature for our support: The customer’s details, including phone number, are automatically saved in HubSpot. If they call us later, staff can immediately see their name and any previously provided information, greatly facilitating communication and increasing the efficiency of customer support.
What were the biggest challenges you faced while building your solution?
I am most proud of the interconnection of the three prompts, which verify each other and try to ensure the most accurate results. The biggest challenge has been processing floor plans, which customers sometimes send in hand-drawn on paper, poorly ochotomised or illegible. This is where the power of our system was demonstrated, as it can extract relevant data even from these poor quality documents.
What are the key modules and functionalities used in your solution?
The key modules in the whole process are undoubtedly Anthropic Claude and working with Google Sheets. Without these technologies, automation and efficient data processing would not be possible.
What are you most proud of about this solution?
I am proud of the entire system connection and the fact that customers praise us. They get a quote immediately without having to wait for anyone, which significantly increases the chance that they will purchase the material immediately.
What feedback have you received from the users of your solution?
The feedback has been fantastic - customers are happy and appreciate the speed and simplicity of the whole process.
What measurable improvements has your solution achieved?
Our designers are no longer burdened with project processing, which allows them to devote more time to individual customer requirements. The time to deliver the offer to the customer has been reduced from approximately 4 days to just minutes.
What makes your solution unique?
I’ve never seen any software that achieves the accuracy in plan recognition that our trained AI does. No competing company has a system this sophisticated and elaborate - let alone one where the customer can generate the quote themselves, without our assistance.
How could this solution be improved?
By improving the prompts and the quality of the AI, but this will take some time.
Is there anything else you would like to share about your solution?
I spent over 100 hours on it, went through a lot of trial and error and solved problems that came up during the testing of the end-to-end steps. Fortunately, it was successful and we are all very happy with it.
Automated communication with hotel guests
This solution was built by @Soichiro_Nishihara and his team.
What specific problem did you aim to solve?
We aimed to solve the overwhelming burden of multilingual customer inquiries faced by Japanese hotels. As international tourism continues to grow and staff shortages worsen, hotel teams are struggling to handle inquiries coming through multiple channels (phone, email, messaging apps, OTA platforms) in various languages. This manual and fragmented approach causes delays, inconsistencies, and high stress levels for staff, ultimately impacting both guest satisfaction and employee retention.
Why was this particular solution necessary?
This solution was necessary because traditional customer support methods were collapsing under unprecedented pressure. Hotels were confronted with a perfect storm: escalating numbers of international guests with diverse language needs, the proliferation of communication channels, and a persistent labor shortage. Relying solely on manual processes led to delayed responses, frequent miscommunication, and increased workload for already overburdened staff. With guest reviews frequently citing frustration at slow or inadequate pre-stay communication, a new, AI-driven approach was essential to maintain service quality without placing additional strain on personnel—especially as Japan sets new records for tourism in the post-pandemic era.
Who is the intended user of this solution?
Our solution is designed for any Japanese hotel or accommodation provider—ranging from boutique inns to large chains—that needs to streamline multilingual guest communications. The primary users are front desk staff, reservation managers, and customer support teams tasked with handling diverse inquiries without sufficient linguistic resources. Secondary beneficiaries include property owners and management companies overseeing multiple properties; they benefit from centralized reporting, consistent service standards, and reduced operational costs. The solution is particularly crucial for smaller or remote properties that struggle to hire multilingual staff.
Provide a brief overview of your solution and how it tackles the problem.
Hosport is a comprehensive AI-powered Business Process Outsourcing (BPO) solution that transforms how hotels handle guest communications. Unlike conventional chatbots, we’ve built an integrated system using Make’s low-code platform to unify fragmented communication channels (phone, email, messaging apps, OTA platforms) into a single interface. The system connects Zendesk for ticket management, Kintone for FAQ knowledge base management, and OpenAI’s Assistant API for intelligent responses, creating a seamless workflow where: Guest inquiries from any channel are captured and consolidated in Zendesk Our AI concierge, trained on hotel-specific information, provides accurate responses in the guest’s native language within approximately 7 seconds For complex queries beyond AI capabilities, staff are instantly notified via Slack (or LINE/ChatWork) for human intervention All interactions are logged, analyzed, and used to continually improve the AI’s knowledge base This solution has delivered remarkable results for our clients, reducing staff workload by up to 3 hours daily while handling 70%+ of inquiries automatically, even with Japanese hotels’ uniquely detailed service requirements. WebPape：https://hos-port.com/
What were the biggest challenges you faced while building your solution?
The most significant challenge we faced was integrating the diverse communication channels used in the hospitality industry. Hotels receive inquiries through multiple platforms (OTA messaging systems, LINE, WhatsApp, telephone, email), each with unique APIs and data structures. We recognized that implementing AI responses on only a few channels would create fragmentation and increase staff confusion rather than reducing workload.
Creating a truly unified communication system required:
Complete Channel Integration: We had to develop custom connectors for each channel, particularly challenging for OTA messaging platforms that lacked standard APIs. This required working with Booking.com, Airbnb, Agoda, and others to establish reliable message capture systems.
Unified Data Model: Each channel formats guest information differently. We developed a standardized data structure that preserves all essential context while creating consistency for AI processing.
Seamless Handoff Protocol: When inquiries exceed AI capabilities, the system must transfer to human staff with full context. This required careful design of notification workflows that provide staff with complete conversation history regardless of the original channel.
Multilingual Voice Integration: Adding telephone support presented unique challenges beyond text-based channels. We implemented a sophisticated system combining voice recognition, AI processing, and text-to-speech capabilities to provide a natural conversation experience.
API Reliability Management: We encountered significant challenges with the OpenAI Assistant API, which returns errors approximately 2% of the time. This required sophisticated error handling within our Make workflows. We implemented loop mechanisms that retry failed API calls, with intelligent timing adjustments for rate limit errors. Through careful engineering of these recovery processes, we successfully achieved zero response failures for hotel guests, ensuring a reliable service despite upstream API inconsistencies.
Our clients particularly value this channel unification and reliability, as it allows their staff to work within a single interface while providing guests with the convenience of using their preferred communication methods, all without experiencing service interruptions.
What are the key modules and functionalities used in your solution?
Our solution leverages several key Make modules and functionalities to create a sophisticated AI-powered communication system:
HTTP/Webhook Modules: These form the foundation of our multi-channel integration, enabling real-time reception and processing of events from Zendesk, messaging platforms (LINE, WhatsApp), email services, and OTA messaging systems. This capability allows us to create a unified communication hub that captures inquiries from every touchpoint.
OpenAI Integration: We extensively utilize Make’s OpenAI modules to connect with the Assistant API. These modules manage conversation context, generate responses, and handle token optimization for cost-effective operation. The integration allows us to process inquiries from all channels through a centralized AI engine that maintains consistent service quality.
FAQ Management Workflow: Our custom Kintone application connected through Make allows hotel staff to easily maintain their knowledge base. When staff click the “learning button,” a Make workflow converts Kintone FAQ data and updates the Assistant API’s knowledge base, giving hotels direct control over AI responses without technical expertise.
Escalation System: We implemented an intelligent workflow using Make’s OpenAI modules to analyze inquiry content and determine when human intervention is required. When escalation criteria return “true,” the system automatically generates detailed notifications in Slack with full conversation context, enabling seamless handoffs.
Twilio Integration: This critical module supports voice call transcription, AI processing of voice inquiries, and text-to-speech response generation for telephone interactions. The integration allows us to extend the same AI capabilities to voice channels that we provide for text-based communications.
Router and Data Transformation Modules: Make’s tools for conditional routing and data transformation are essential for normalizing inputs from diverse channels into a consistent format for AI processing, and then adapting responses to meet the requirements of each communication platform.
The true innovation lies in how Make orchestrates these components into a cohesive system that operates seamlessly across all guest communication touchpoints while maintaining a unified conversation context.
What are you most proud of about this solution?
What we’re most proud of about Hosport is how it transforms the fundamental economics of hospitality operations while simultaneously enhancing the guest experience. Our solution doesn’t merely automate responses—it completely reimagines the guest communication workflow by:
Delivering Impossible Metrics: Achieving an average response time of 7 seconds in any language, 24/7/365—a standard that would be financially impossible to maintain with human-only staff.
Bridging Technology and Tradition: Successfully applying cutting-edge AI to an industry known for its emphasis on human touch, without compromising the exceptional service standards synonymous with Japanese hospitality.
Creating Measurable Impact: Generating tangible results for our clients—reducing staff workload by up to 3 hours daily, handling over 70% of inquiries automatically, and significantly improving guest satisfaction metrics.
Building a Learning System: Developing an AI that continuously improves through every interaction and every manual staff intervention, creating a virtuous cycle of ever-improving service quality.
The real achievement is the delicate balance we’ve struck: using Make to orchestrate a complex system that feels completely invisible and seamless to both hotel staff and guests. Hotels can now focus resources on providing extraordinary in-person experiences, while we handle the digital communication layer with equal attention to detail and service quality. In an industry struggling with staffing shortages, this solution doesn’t replace hospitality workers—it empowers them to deliver better service by focusing on what humans do best.
What feedback have you received from the users of your solution?
The feedback from hotels using our solution has been overwhelmingly positive, particularly highlighting the operational improvements and their impact on service quality:
As shown in the testimonial from the UNPLAN FUKUOKA hotel manager, staff previously spent 3-4 hours daily managing communications across email, LINE, and OTA messages. After implementing Hosport, they now spend only about 30 minutes checking that everything is running smoothly. The manager specifically notes: “The freed-up time can now be used for in-person guest interactions, event planning, and local information gathering, which has greatly improved our word-of-mouth reputation.”
Hotel operators consistently emphasize the value of channel integration. As one manager stated, “While the AI responses are certainly important, the greatest impact comes from having all messaging channels unified in one system.”
Staff members have also developed a clearer understanding of their unique value proposition in the guest experience. According to the UNPLAN FUKUOKA manager: “Because AI can handle so many routine inquiries effectively, we’ve become more conscious about ensuring we deliver value that only humans can provide, and we’re communicating this philosophy to all our staff.”
This feedback confirms that Hosport is achieving its core mission: freeing hotel staff from routine communications so they can focus on creating exceptional guest experiences.
What measurable improvements has your solution achieved?
Our solution has delivered concrete, measurable improvements for hotel operations:
Operational Efficiency: As evidenced by the UNPLAN FUKUOKA implementation, staff time spent on communication tasks has been reduced from 3-4 hours daily to approximately 30 minutes—a reduction of 83-88% in manual communication workload.
AI Resolution Rate: From August 1 to October 9, our system successfully resolved 72% of all inquiries without requiring staff intervention. The remaining 28% of inquiries were seamlessly escalated to staff, ensuring no guest request went unaddressed.
Response Time: Average response time to guest inquiries decreased from hours to just 7 seconds, representing a near-immediate response capability regardless of time of day or language.
Staff Reallocation: The time freed up has allowed hotels to reallocate staff resources to high-value activities such as in-person guest service, event planning, and local information gathering—activities that directly enhance the guest experience and drive revenue.
Guest Satisfaction: Properties using our solution have seen measurable improvements in review scores specific to communication and pre-arrival experience.
These metrics demonstrate that Hosport delivers transformative operational improvements while enhancing the guest experience, creating a virtuous cycle of better service with lower operational burden.
What makes your solution unique?
What sets Hosport apart from other AI solutions in the hospitality industry is our comprehensive, end-to-end approach that addresses the complete guest communication lifecycle:
True Omnichannel Integration: Unlike competitors who focus on single channels (typically chatbots), Hosport unifies all communication touchpoints—including voice calls, a critical channel that most AI solutions ignore. This complete integration ensures a consistent guest experience regardless of communication method.
Built for Hospitality: Our solution is purpose-built for the unique requirements of Japanese hospitality, with its emphasis on meticulous attention to detail and exceptional service standards. The system understands hospitality-specific context and nuances that general-purpose AI solutions miss.
Human-AI Collaboration Model: Rather than attempting to replace staff, we’ve designed a system that handles routine inquiries while seamlessly transitioning to human assistance when needed. Our solution acts as an intelligent assistant that amplifies staff capabilities rather than attempting to replace them.
Custom Knowledge Base Architecture: Each property has unique facilities, policies, and surroundings. Our Kintone-based knowledge management system allows non-technical staff to easily maintain and update the AI’s information, ensuring responses remain accurate and property-specific.
Workflow Automation Through Make: By leveraging Make’s low-code platform, we’ve created a highly adaptable system that can be customized to each property’s specific processes and integrated with their existing technology stack, rather than forcing them to adapt to our system.
This combination of hospitality expertise, technology integration, and collaborative design philosophy produces a solution that addresses the full complexity of hotel operations rather than merely applying AI to isolated communication channels.
How could this solution be improved?
While our solution delivers substantial value, we’ve identified several areas for continued enhancement:
Deeper Integration with Property Management Systems: Currently, our solution operates with limited access to reservation data. By developing deeper integrations with PMS platforms like Temairazu, Neppan, and Beds24, we could provide more personalized responses based on guest history, preferences, and upcoming stay details.
Enhanced Predictive Analytics: With the growing dataset of guest inquiries and responses, we can develop predictive models that anticipate common questions based on factors like time before arrival, guest demographics, and booking channel. This would allow for proactive communication that addresses potential concerns before they’re even asked.
Advanced Voice AI Capabilities: While our voice integration is functional, we aim to improve the naturalness of voice responses and add capabilities like sentiment analysis that can detect caller emotions and adjust response tone accordingly.
Expanded Revenue Generation Opportunities: The system could be enhanced to identify and act on upselling opportunities during guest interactions, suggesting room upgrades, ancillary services, or special packages when appropriate.
Automated Quality Assurance: Implementing ongoing automated analysis of AI responses to ensure they maintain the property’s service standards and detect any potential degradation in response quality before it impacts guests.
We’re actively working on these improvements as part of our development roadmap, with PMS integrations as our highest priority based on client feedback.
Is there anything else you would like to share about your solution?
An often-overlooked aspect of our solution is its positive impact on hotel staff satisfaction and retention. In an industry plagued by high turnover rates, particularly in Japan where the labor shortage is acute, Hosport provides significant benefits beyond operational efficiency.
Front desk staff consistently report higher job satisfaction after implementation, as they’re freed from repetitive inquiries and can focus on meaningful guest interactions. This shift from reactive, routine response work to proactive, high-value guest engagement transforms the nature of their roles in positive ways.
One hotel manager shared that before Hosport, new staff members would spend their first few months memorizing standard responses to common questions, creating a tedious onboarding experience. With Hosport handling these routine inquiries, new employees now engage in more fulfilling guest service activities from day one, accelerating their development and improving retention.
From a technical perspective, our use of Make has been transformative in allowing us to rapidly iterate and improve the solution based on client feedback. The low-code environment enables us to implement enhancements in days rather than weeks, responding to the unique needs of each property while maintaining a robust core system. This adaptive capability has been crucial to our success in an industry where each property has distinct requirements and service philosophies.
Finally, we believe Hosport represents the future of hospitality—not by replacing the human element that makes hospitality special, but by enhancing it through technology that handles routine tasks while enabling staff to focus on creating the meaningful connections and exceptional experiences that guests remember.
