Hotel Monitoring and Management Software NEXT SMS System Development Agreement
Exhibit 10.2
Hotel Monitoring and Management Software NEXT
SMS System
Development Agreement
Date of Signing: 2025, June 27
Parties to the Agreement
Party A (Client): ALOHA Asia Pacific Limited
Party B (Developer): Next Technology Holding Inc.
Whereas:
| 1. | Party A is a company established and existing in accordance with the relevant laws and regulations of Hong Kong, with full legal personality. It has obtained authorization through a board resolution to purchase the AI+SaaS technical support system NEXT SMS pursuant to this Agreement, and operates 20 hotels and resorts in the Asia-Pacific region; |
| 2. | Party B is a limited liability company established and existing in accordance with the registration requirements of the U.S. Securities and Exchange Commission (SEC) and the Securities Exchange Act of 1934. Its common stocks are publicly traded on the NASDAQ Stock Exchange (NASDAQ ticker symbol: NXTT). With complete legal status, Party B is deeply engaged in the field of artificial intelligence + software-as-a-service development, and has profound technical accumulation especially in intelligent security management systems, AI model quantization and deployment, federated learning, real-time AI inference engine development, data encryption, SaaS multi-tenant architecture design, and edge AI computing. It also holds authoritative technical certifications such as ISO 27001, AWS AI/ML Certification, and Microsoft Azure SaaS Solution Certification; |
| 3. | Due to business and technical upgrading needs, Party A urgently needs to build the NEXT SMS System (AI+SaaS technical support system, hereinafter referred to as the “NEXT SMS System”) centered on hotel security management services (integrating AI-driven intelligent analysis, edge-cloud collaborative computing, and cloud-based SaaS service capabilities) to improve operational efficiency, strengthen hotel security monitoring capabilities, and reduce potential security risks. Party B undertakes to provide full-cycle consulting services, including demand diagnosis, AI model customization and fine-tuning, system integration, AI model compression and optimization, performance tuning, and continuous maintenance; |
1. Definitions and Interpretations
1.1 “NEXT SMS System” (Hotel Monitoring and Management Software AI+SaaS Platform, hereinafter referred to as the “NEXT SMS System”) refers to the cloud-based software service platform developed by Party B for Party A. This platform integrates AI-driven core functions (such as anti-counterfeiting face recognition, behavior recognition, risk prediction model, edge AI inference, and anomaly detection algorithm) and SaaS basic functions (real-time video monitoring, motion detection, alarm management, encrypted storage, remote access, and intelligent report generation). Adopting a SaaS multi-tenant architecture, the platform can uniformly manage 20 hotels/resorts (each with independent data isolation and permission control) and enable convenient access through web browsers and mobile applications.
1 / 18
1.2 “Development Cycle” refers to an 18-month period from the effective date of this Agreement to the final acceptance date, divided into three phases: Phase 1 (Demand Analysis and AI Model Architecture Design), Phase 2 (AI Algorithm Development and SaaS Multi-Tenant System Testing), and Phase 3 (Edge AI Gateway Deployment and AI Function Training of the NEXT SMS System).
1.3 “Acceptance Criteria” refers to the various technical standards and functional standards specified in Appendix I, including but not limited to AI model performance indicators (precision, recall rate, F1 score), SaaS multi-tenant isolation effect, stability of the NEXT SMS System, response speed, compliance with Hong Kong PDPO (Personal Data (Privacy) Ordinance), and compatibility with third-party devices.
1.4 “Service Level Agreement (SLA)” refers to Party B’s commitment to ensure the availability of the NEXT SMS System after launch (≥99.9%), AI inference service throughput (≥500 TPS per hotel), AI model accuracy retention rate (≥95% within 6 months), fault response time (≤4 hours), and fault resolution time (≤24 hours), as described in Article 6.
1.5 “AI Model” refers to the machine learning models developed by Party B for the NEXT SMS System, including but not limited to deep learning-based face recognition models, Convolutional Neural Network (CNN) behavior classification models, Long Short-Term Memory (LSTM) risk prediction models, and lightweight TensorRT-optimized edge inference models.
1.6 “Edge AI Inference” refers to the technical capability of deploying quantized AI models on local monitoring devices (such as cameras, NVRs, and edge AI gateways) to realize real-time analysis of video streams without relying on cloud computing resources, thereby reducing the end-to-end latency of the NEXT SMS System.
1.7 “SaaS Multi-Tenant Architecture” refers to a cloud service architecture in which a single instance of the NEXT SMS System serves 20 hotels/resorts (tenants), logically isolates tenant data (through data partitioning and permission mapping), shares hardware/software resources, and supports personalized configuration for each hotel.
1.8 “Federated Learning” refers to a distributed AI training technology where edge nodes of each hotel use their own anonymized video data to train AI models locally, and only upload model parameters (not raw data) to the cloud for aggregation. This ensures data privacy while improving the cross-scenario AI generalization capability of the NEXT SMS System.
1.9 “Edge AI Gateway” refers to the hardware-software integrated component deployed in each hotel, which is responsible for edge AI inference scheduling, video data preprocessing (cropping, compression), encrypted transmission between edge devices and the cloud, and local fault recovery of the AI services of the NEXT SMS System.
1.10 “AI Model Quantization” refers to the technical process of converting a 32-bit floating-point AI model into an 8-bit integer model to reduce the model size (≥70%) and inference latency (≥50%), adapt to the limited computing resources of edge devices, and ensure the operating efficiency of the NEXT SMS System at the edge.
2 / 18
2. Scope of Development Services
Party B shall provide the following development services for the NEXT SMS System (AI+SaaS), ensuring compliance with industry best practices (such as ISO/IEC 25010 Software Quality Standard, IEEE 802.11 Security Protocol), Hong Kong Security and Guarding Services Ordinance, and Party A’s operational requirements for 20 hotels/resorts:
2.1 Architecture Design of the NEXT SMS System (AI+SaaS)
| ● | Design a cloud-edge collaborative multi-tenant architecture to support ≥500 users accessing the NEXT SMS System simultaneously across 20 hotels/resorts: Deploy a central cloud SaaS platform for unified management, and configure 1 edge AI gateway for each hotel to run quantized AI models (reducing cloud bandwidth usage by ≥40% and AI inference latency by ≥60%); |
| ● | Develop dedicated API interfaces for integration with Party A’s existing PMS (Property Management System), access control system, and fire alarm system; design an AI model parameter synchronization protocol (to ensure real-time alignment of cloud/edge AI models and avoid data consistency issues of the NEXT SMS System); |
| ● | Ensure the compatibility of the NEXT SMS System with mainstream surveillance cameras (such as Hikvision, Dahua) and edge AI gateways that support ONVIF/GB28181 protocols, and optimize the deployment of AI models based on TensorRT/TFLite for different hardware configurations (CPU/GPU/NPU) of edge devices. |
2.2 Core Function Development of the NEXT SMS System (AI+SaaS)
| ● | Real-time AI Monitoring Module: Supports multi-camera real-time monitoring, equipped with AI-driven PTZ automatic tracking function (locks and follows abnormal targets through target detection algorithms), and video recording using AI dynamic compression technology (adjusts bit rate according to motion intensity, saving storage space by ≥30%). This module integrates a real-time AI inference engine, which can run quantized behavior recognition models locally to achieve inference latency of ≤100 ms, ensuring the real-time performance of the NEXT SMS System; |
| ● | Intelligent AI Alarm System: Equipped with AI-driven motion detection technology (with a false alarm rate of ≤3% through background modeling and noise reduction), AI-enhanced face recognition function (accuracy ≥98% in low-light/backlight environments, supporting 3D anti-counterfeiting using depth camera data), AI anomaly detection technology (identifies illegal intrusion, loitering, lost items, and elevator overloading based on CNN-LSTM hybrid model), and AI priority alarm routing system (classifies alarms into high/medium/low risk levels through risk prediction models and pushes them to designated personnel via SMS/email/APP), forming the core security capability of the NEXT SMS System; |
3 / 18
| ● | AI Data Analysis Dashboard: Generates daily/weekly descriptive analysis reports (alarm frequency, camera downtime, staff patrol records) and predictive analysis reports (high-risk area early warning, peak alarm time prediction) through data mining algorithms. The dashboard also includes an AI model performance monitoring module, which displays real-time performance indicators such as accuracy, recall rate, and inference latency of edge AI models in each hotel, providing data support for the operation and maintenance of the NEXT SMS System; |
| ● | SaaS Multi-Tenant User Management: Supports Role-Based Access Control (RBAC) function to achieve tenant-level permission isolation — hotel administrators can only manage the AI+SaaS functional modules of the NEXT SMS System for their own hotels, while Party A’s headquarters has cross-tenant supervision rights. This module integrates AI-assisted operation audit function (identifies unauthorized permission changes and abnormal login behaviors through anomaly detection algorithms) and retains 12-month operation trajectory records, complying with the compliance requirements of the NEXT SMS System. |
2.3 Security and Compliance Functions of the NEXT SMS System (AI+SaaS)
| ● | Encrypt the video data of the NEXT SMS System during transmission (TLS 1.3) and at rest (AES-256), and implement AI-based dynamic data anonymization in accordance with PDPO (automatically mask guests’ faces, license plates, and room numbers in the camera using semantic segmentation models); |
| ● | Adopt federated learning for AI model training of the NEXT SMS System — raw video data is retained in each hotel, and only model parameters are uploaded to the cloud to avoid cross-hotel data leakage and meet the data localization requirements of PDPO; |
| ● | Use SHAP/LIME algorithms to provide AI Model Explainability (XAI) reports for key decisions of the NEXT SMS System (e.g., reasons for face recognition rejection, triggering of abnormal alarms), ensuring compliance with audit requirements for AI-driven monitoring; |
| ● | Establish a data lifecycle management system for the NEXT SMS System in compliance with GDPR/PDPO — AI analysis results are retained for 1 year, raw video data is retained for 30 days, and expired data is automatically deleted through an AI-driven data cleaning strategy (prioritizing the deletion of non-critical footage). |
2.4 Testing and Optimization of the NEXT SMS System (AI+SaaS)
| ● | Conduct SaaS multi-tenant load testing: Verify whether the NEXT SMS System can support 20 concurrent tenants, 1000+ simultaneous video streams, and 500+ AI inference requests per second (TPS) without performance degradation; |
| ● | Conduct AI model verification testing: Entrust a third-party testing company to evaluate the AI model performance of the NEXT SMS System using confusion matrix analysis, ROC curve verification, and adversarial attack resistance testing (ensuring face recognition precision ≥97%, recall rate ≥96%, and behavior recognition ≥95%); |
4 / 18
| ● | Optimize edge-cloud collaboration performance: Adjust the data preprocessing logic of the Edge AI Gateway (cropping irrelevant video areas) and AI model quantization parameters (balancing precision and latency) to reduce the end-to-end video transmission + AI inference latency of the NEXT SMS System to ≤200 ms; |
| ● | Conduct SaaS tenant isolation testing: Verify that data/configuration changes in the tenant space of one hotel do not affect other tenants, and that the NEXT SMS System blocks unauthorized cross-tenant access attempts. |
2.5 Training and Documentation for the NEXT SMS System (AI+SaaS)
| ● | Provide on-site training (≥5 sessions) for Party A’s employees, including SaaS multi-tenant operation training (tenant configuration, cross-hotel data viewing), AI model maintenance training (parameter adjustment, quantized model update, federated learning node management), and AI troubleshooting training (inference failure recovery, model rollback), ensuring that Party A can proficiently operate the NEXT SMS System; |
| ● | Deliver technical documents of the NEXT SMS System, including user manual, API specifications (tenant-specific interfaces), AI model documents (algorithm principles, training datasets, quantization parameters, TensorRT optimization logs), SaaS multi-tenant architecture diagram, Edge AI Gateway deployment guide, and federated learning operation manual. |
3. Development Schedule and Deliverables
3.1 Phase 1 (Months 1-3): Demand Analysis and AI Model Architecture Design
| ● | Deliverables: Detailed demand specifications (including tenant-specific AI function requirements of the NEXT SMS System for 20 hotels/resorts), multi-tenant system architecture design document of the NEXT SMS System (cloud-edge collaboration, data isolation solution), AI model architecture plan (algorithm selection, federated learning node design, quantization plan), and UI/UX prototype (tenant-specific dashboard). |
| ● | Milestone: Approval by Party A’s technical team. |
3.2 Phase 2 (Months 4-15): AI Algorithm Development and SaaS Multi-Tenant System Testing
| ● | Deliverables: Alpha/Beta versions of the NEXT SMS System (including quantized edge AI models, SaaS multi-tenant core modules), test reports (AI model performance test results of the NEXT SMS System, SaaS multi-tenant load test records, security penetration test reports), bug fix logs, and AI model optimization plans (adjustment plans for low recall rates in specific hotel scenarios). | |
| ● | Milestone: Successful completion of Beta testing, with ≥95% of the core functions of the NEXT SMS System meeting the acceptance criteria (approved by Party A). |
5 / 18
3.3 Phase 3 (Months 16-18): Edge AI Gateway Deployment and AI Function Training of the NEXT SMS System
| ● | Deliverables: Production-ready NEXT SMS System (deployed to 20 hotels/resorts with SaaS multi-tenant function enabled), Edge AI Gateway configuration manual (provided per hotel), training materials (AI model maintenance of the NEXT SMS System, SaaS tenant management), and User Acceptance Testing (UAT) report (covering tenant isolation verification, AI inference latency testing, and federated learning function verification). |
| ● | Milestone: Final acceptance by Party A. |
4. Development Fees and Payment Method
4.1 Total Development Fee: USD 4,220,000 (excluding taxes). This fee is calculated based on 20 hotels/resorts operated by Party A, with a unit price of USD 211,000 per hotel/resort (including AI model customization and quantization of the NEXT SMS System, SaaS multi-tenant development, Edge AI Gateway deployment, and federated learning node configuration).
4.2 Payment Schedule:
| ● | 50% (USD 2110000) shall be paid within 10 working days after signing this agreement as a prepayment for the AI model architecture design and SaaS multi tenant framework construction of the NEXT SMS system; |
| ● | 50% (USD 2,110,000) shall be paid upon final acceptance (issued by Party A) after the NEXT SMS System is fully deployed to 20 hotels/resorts, the SaaS multi-tenant isolation verification is completed, and the AI model performance meets the acceptance criteria. |
4.3 Payment Method: Bank transfer to the designated account of Party B (Next Technology Holding Inc.).
5. Rights and Obligations
5.1 Rights and Obligations of Party A
| ● | Provide clear requirements for the hotel-specific functions of the NEXT SMS System (such as AI behavior recognition scenarios exclusive to resort hotels) and feedback; within 7 days after Party B’s request, grant access rights to relevant systems/devices (including edge cameras and NVRs) of 20 hotels/resorts to ensure the smooth progress of the NEXT SMS System development; |
| ● | Appoint a full-time project coordinator to communicate with Party B, and approve the deliverables (such as the AI model architecture plan of the NEXT SMS System, SaaS multi-tenant design) within 10 days after submission; |
6 / 18
| ● | Cooperate with Party B in preparing data for federated learning — provide anonymized historical video data for each hotel within 15 days after the launch of Phase II, ensuring compliance with PDPO and meeting the AI model training needs of the NEXT SMS System; |
| ● | Make payments in accordance with Article 4; if the payment is overdue for more than 15 days, a late fee of 0.05% of the total overdue amount shall be charged daily. |
5.2 Rights and Obligations of Party B
| ● | Complete the development task of the NEXT SMS System in accordance with the schedule specified in Article 3, ensuring that the SaaS multi-tenant isolation effect (100% data separation between tenants) and AI model performance (meeting precision/recall indicators) comply with the acceptance criteria (Appendix I); |
| ● | Within 3 days after discovering risks (such as insufficient accuracy of the AI model in the low-light environment of the hotel lobby, SaaS multi-tenant concurrency bottlenecks of the NEXT SMS System), notify Party A in writing and propose feasible mitigation plans (such as fine-tuning the AI model using hotel-specific data, optimizing SaaS load balancing, etc.); |
| ● | Provide 12 months of post-launch technical support in accordance with the SLA (Annex II), including quarterly AI model retraining (through federated learning), SaaS multi-tenant vulnerability repair of the NEXT SMS System, and Edge AI Gateway firmware update; |
| ● | Ensure that the NEXT SMS System does not infringe any third party’s intellectual property rights — especially AI algorithms, quantized model weights, SaaS multi-tenant core codes, etc. In case of infringement claims, Party B shall bear full responsibility (including compensating for the losses suffered by Party A). |
6. Intellectual Property Rights
6.1 After the full payment of the total development fee (as described in Article 4.1), all intellectual property rights of the NEXT SMS System (including SaaS multi-tenant source code, AI models (quantized/non-quantized), federated learning algorithms, Edge AI Gateway configuration scripts, design documents, AI training datasets (anonymized), and test reports) shall be fully owned by Party A (ALOHA Asia Pacific Limited).
6.2 The Party B has the right to repurchase intellectual property rights: After the Party A obtains exclusive intellectual property rights for the NEXT SMS system in accordance with Article 6.1, Party B (Next Technology Holding Inc.) has the right to repurchase the system’s intellectual property rights (excluding Party A’s tenant exclusive configuration data and business data) within three years. The repurchase price is [] US dollars , and the specific amount will be jointly confirmed by both Parties based on the market value of intellectual property at that time, including the estimation of the NEXT SMS system AI model
| ● | With the written consent of Party A, Party B shall pay the full repurchase price within 30 days; after the payment is confirmed, the relevant intellectual property rights shall be transferred to Party B, and both parties shall sign a supplementary agreement to confirm the scope of transfer (e.g., excluding the tenant isolation module) and the obligations after transfer (e.g., Party B shall not use the repurchased AI model for hotels in Party A’s direct market). |
7 / 18
| ● | Within two years after the repurchase, Party B shall not use the repurchased intellectual property rights to develop competitive NEXT SMS System (AI+SaaS)-type products for Party A’s direct competitors in the Asia-Pacific hotel industry. |
6.3 Party B reserves the right to use technical frameworks and methods not dedicated to this project (such as public AI development frameworks like TensorFlow/PyTorch, SaaS multi-tenant open-source templates) for other projects, but shall not disclose Party A’s confidential information (such as tenant data isolation logic, AI model parameters of the hotel-specific NEXT SMS System).
7. Confidentiality
7.1 Both Party A and Party B shall keep confidential all information obtained during the cooperation period (including Party A’s hotel operation data, tenant-specific AI requirements for the NEXT SMS System, Party B’s AI algorithm details, SaaS multi-tenant core codes, federated learning parameter aggregation logic, test data, etc.) for five years from the date of termination of this Agreement.
7.2 If either party breaches the confidentiality obligation, the non-breaching party shall have the right to claim compensation for losses, including reasonable legal expenses incurred for safeguarding its rights (e.g., expenses for investigating the leakage of the AI algorithm of the NEXT SMS System).
8. Acceptance
8.1 UAT Process: Party A shall conduct UAT within 30 days after receiving the final version of the NEXT SMS System delivered by Party B. The test shall cover all AI and SaaS functions listed in Appendix I, including:
| ● | SaaS Multi-Tenant Isolation Test: Verify that data/configuration changes in 3 randomly selected hotels do not affect other tenants; |
| ● | Random Check of AI Model Performance: Test face recognition accuracy, behavior recognition recall rate, and inference latency in 5 hotels (including 2 resort hotels with complex outdoor scenarios); |
| ● | Federated Learning Function Test: Verify that the uploaded model parameters do not include raw data, and the aggregated model can improve the performance of the NEXT SMS System in all hotels. |
8.2 Acceptance Results:
| ● | Pass: Party A issues a final acceptance certificate to Party B, confirming that the NEXT SMS System meets the requirements of all 20 hotels/resorts. |
8 / 18
| ● | Conditional Pass: Party A notifies Party B of defects (such as the AI model recall rate of the NEXT SMS System in the resort garden ≤95%, SaaS tenant login latency > 3s), and Party B shall rectify the defects within 15 days (such as fine-tuning the AI model using outdoor scenario data, adjusting SaaS load balancing). After rectification, Party A shall conduct re-acceptance. |
| ● | Fail: Party B shall redevelop the defective modules (including AI model retraining, optimization of the SaaS multi-tenant module of the NEXT SMS System) within 30 days without charging any additional fees to Party A, and submit the redeveloped modules to Party A for acceptance again. |
9. Termination
9.1 If one party materially breaches any clause of this Agreement and fails to rectify the breach within 30 days after receiving a written notice from the non-breaching party (e.g., Party B fails to achieve SaaS multi-tenant isolation of the NEXT SMS System after two rectifications, or Party A delays payment for more than 45 days), the non-breaching party may terminate this Agreement.
9.2 After the termination of this Agreement, Party B shall deliver all completed work results (including quantized AI models, SaaS multi-tenant source code of the NEXT SMS System, federated learning logs, test reports) and relevant materials to Party A. Party A shall make proportional payment based on the workload completed by Party B (confirmed by both parties; for example, if Phase II work is completed and the AI model performance reaches 90% of the acceptance criteria, 50% of the payment shall be made).
10. Dispute Resolution
Any dispute arising from or in connection with this Agreement shall first be resolved by the parties through friendly negotiation; if negotiation fails, the dispute shall be submitted to the Hong Kong International Arbitration Centre (HKIAC) for arbitration in accordance with its currently effective arbitration rules. The arbitration award shall be final and binding on both parties, and shall cover matters such as liability for breach of AI model performance of the NEXT SMS System and compensation for SaaS multi-tenant defects.
11. Miscellaneous
11.1 This Agreement shall take effect from the date of signature by the authorized representatives of both parties and shall remain valid until all obligations of both parties under this Agreement are fully performed.
11.2 This Agreement and its appendices (Appendix I: Acceptance Criteria; Annex II: Service Level Agreement) constitute the complete agreement reached by both parties on the subject matter of this Agreement (development of the NEXT SMS System AI+SaaS), and supersede all previous oral or written agreements, understandings, and communications between the parties (including discussions on AI model selection and SaaS architecture).
11.3 Any modification or supplement to this Agreement shall be made in writing and shall take effect only after being signed by the authorized representatives of both parties (e.g., adjustment of the AI model performance indicators of the NEXT SMS System or the number of SaaS multi-tenants).
[Signature Page Follows]
9 / 18
Party A (Client): ALOHA Asia Pacific Limited
Authorized Signature: /s/ ALOHA Asia Pacific Limited
Party B (Developer): Next Technology Holding Inc.
Authorized Signature: /s/ Next Technology Holding Inc.
10 / 18
Appendix I: Acceptance Criteria (NEXT SMS System)
Acceptability Checklist
Basic Project Information
| Item | Details |
| Project Name | Hotel Monitoring and Management Software NEXT SMS System |
| Acceptance Date | ____________________ |
| Party A (Client) | ALOHA Asia Pacific Limited |
| Party B (Developer) | Next Technology Holding Inc. |
1. Core Acceptance Items (Mark “√” for Pass, “×” for Fail)
| No. | Acceptance Category | Specific Inspection Item | Technical Requirement | Pass (√) | Fail (×) | Remarks |
| 1 | Completeness of Deliverables | 1.1 Technical Documents | Including cloud-edge multi-tenant architecture document, tenant API specifications, AI model document of the NEXT SMS System, Edge AI Gateway manual | |||
| 1.2 SaaS Multi-Tenant Source Code + Quantized AI Models (TensorRT/TFLite) |
11 / 18
| 1.3 Test Report (Function Coverage ≥95%, Motion Detection False Alarm Rate ≤3%) | ||||||
| 1.4 ISO 27001 Certification + AI Security Report of the NEXT SMS System + ≥5 Training Records | ||||||
| 2 | AI Model Performance | 2.1 Face Recognition | Accuracy ≥98% (Low-light/Backlight) | |||
| 2.2 Behavior Recognition (Trespassing/Loitering) | Precision ≥97%, Recall Rate ≥96% | |||||
| 2.3 Risk Prediction Model | Accuracy ≥92% | |||||
| 3 | System Latency and Concurrency | 3.1 Real-time Video Latency | ≤200 ms (supports 4K) | |||
| 3.2 Edge AI Inference Latency | ≤100 ms (Face/Behavior Recognition) | |||||
| 3.3 Concurrent Capacity | Supports 20 hotels, ≥500 users; AI Throughput ≥500 TPS per hotel |
12 / 18
| 4 | Data Security and Compliance | 4.1 Data Encryption | TLS 1.3 (Transmission) + AES-256 (Storage); No Leakage | |||
| 4.2 PDPO/GDPR Compliance | AI Dynamic Anonymization, Federated Learning (Hotel Raw Data) | |||||
| 4.3 IP Compliance | AI Algorithm License/Originality of the NEXT SMS System; Open Source ≤10% | |||||
| 5 | System Availability | 5.1 Platform Availability | ≥99.9% (30-day Trial) | |||
| 5.2 Edge AI Gateway Availability | ≥99.95% | |||||
| 5.3 Fault Recovery | AI Faults of the NEXT SMS System Repaired ≤12h; General Faults Repaired ≤24h | |||||
| 6 | Hardware Compatibility | 6.1 Cameras and Gateways | Supports Hikvision/Dahua (ONVIF/GB28181), NVIDIA Jetson | |||
| 6.2 Multi-port Operating Systems | Compatible with Windows/iOS/Android |
2. Acceptance Process Verification
| Process Phase | Required Operation | Completion Time (√/×) | Completion Date |
| Phase 1: Application Submission | Party B submitted AI test video acceptance documents of the NEXT SMS System for 3 hotels | ||
| Phase 2: Preliminary Review | Party A verified SaaS tenant isolation + completeness of AI documents of the NEXT SMS System | ||
| Phase 3: Expert Review | Blind AI Test + SaaS Stress Test Completed | ||
| Phase 4: Final Decision | Acceptance Conclusion Confirmed by Both Parties |
13 / 18
3. Final Acceptance Conclusion
☐ Full Pass (All items in Table 1 are marked “√“)
☐ Conditional Pass (≤3 non-core items marked “×”, requiring rectification: ____________________)
☐ Fail (≥1 core item marked “×”, e.g., data leakage, AI accuracy of the NEXT SMS System < 95%)
Rectification Deadline (if applicable): ____________________ (DD/MM/YYYY)
4. Signature Confirmation
Both parties confirm the above acceptance results and agree to implement the corresponding follow-up actions.
Party A (Client): ALOHA Asia Pacific Limited
Authorized Representative Signature: ____________________
Name: ____________________
Date: ____________________
Seal (if applicable): ____________________
Party B (Developer): Next Technology Holding Inc.
Authorized Representative Signature: ____________________
Name: ____________________
Date: ____________________
Seal (if applicable): ____________________
14 / 18
Annex II: Service Level Agreement (SLA for NEXT SMS System (AI+SaaS))
| 1. | Platform Availability: The availability of the NEXT SMS System shall be ≥99.9%, the availability of the Edge AI Gateway shall be ≥99.95%, and the availability of the AI inference service shall be ≥99.95%. |
| 2. | AI Model Performance Guarantee: The accuracy of the AI model of the NEXT SMS System shall remain ≥95% within 6 months (the accuracy decrease shall not exceed 5%), and the behavior recognition recall rate shall remain ≥94%. |
| 3. | Fault Response and Resolution: |
| ◦ | General Faults (SaaS tenant configuration issues of the NEXT SMS System): Response ≤4 hours, Resolution ≤24 hours; |
| ◦ | AI-related Faults (model inference errors, Edge AI Gateway faults): Response ≤2 hours, Resolution ≤12 hours; |
| ◦ | SaaS Multi-Tenant Core Faults (data isolation faults): Response ≤1 hour, Resolution ≤18 hours. |
| 1. | AI Model Maintenance: Quarterly federated learning model update (aggregating new data from 20 hotels), semi-annual AI model quantization optimization of the NEXT SMS System (reducing latency by ≥10% each time), and annual emergency model retraining (for major scenario changes). |
| 2. | Penalties for Breaching SLA: |
| ◦ | For each 0.1% decrease in the platform availability of the NEXT SMS System below 99.9%, Party B shall deduct 1% of the final payment; |
| ◦ | If the accuracy of the AI model of the NEXT SMS System drops below 95% within 6 months, Party B shall provide one free emergency retraining and deduct 2% of the final payment; |
| ◦ | For delayed fault resolution (exceeding the SLA limit), Party B shall deduct 0.5% of the final payment for each 24-hour delay. |
15 / 18
Hotel Monitoring and Management Software NEXT SMS System Maintenance Agreement
Parties:
| ● | Party A (Service Recipient): ALOHA Asia Pacific Limited |
| ● | Party B (Service Provider): Next Technology Holding Inc. |
Date of Signing: Same as the Main Agreement
Whereas:
Party A and Party B have signed the “Hotel Monitoring and Management Software NEXT SMS System Development Agreement” (hereinafter referred to as the “Main Agreement”) regarding the development of the NEXT SMS System for 20 hotels/resorts under Party A. To ensure the long-term stable operation of the NEXT SMS System — including the maintenance of SaaS multi-tenant performance and the continuous optimization of AI models — both parties agree to sign this “Supplementary Maintenance Agreement” to clarify the specific content of the maintenance services.
1. Scope of Maintenance Services
Party B shall provide annual maintenance services for the NEXT SMS System deployed in 20 hotels/resorts of Party A, including but not limited to:
1.1 SaaS Multi-Tenant System Maintenance: Daily operation monitoring (tenant data isolation, concurrent access stability), vulnerability repair (e.g., tenant login failure), software version upgrade (compatibility with new PMS systems), tenant-specific configuration adjustment (e.g., adding new user roles for resort hotels), to ensure the normal operation of the multi-tenant function of the NEXT SMS System.
1.2 AI Model Maintenance: Quarterly AI model performance evaluation (using confusion matrix, F1 score), parameter adjustment for drift (e.g., accuracy decrease due to new camera angles), semi-annual federated learning retraining (using anonymized data from 20 hotels), and optimization of quantized models (adapting to hardware upgrades of Edge AI Gateways), to ensure that the AI model performance of the NEXT SMS System meets the standards.
1.3 Edge AI Gateway Maintenance: Monthly firmware update, debugging of edge AI inference engine (e.g., TensorRT runtime errors), local data backup verification (ensuring no loss of AI analysis results of the NEXT SMS System).
1.4 Technical Support: Provide 7×24 remote technical support (phone/email) for general issues, and on-site support within 48 hours for emergency faults (e.g., hardware faults of Edge AI Gateways, abnormal SaaS multi-tenant data isolation of the NEXT SMS System).
1.5 Training and Reports: Provide one free training session per hotel per year (covering AI model performance monitoring of the NEXT SMS System and operation of new SaaS multi- tenant functions), and a monthly maintenance report (including fault records, AI model accuracy trends, and Edge AI Gateway uptime).
16 / 18
2. Maintenance Term
3 years after the final acceptance date of the NEXT SMS System (automatically extended if there is no written notice from both parties).
3. Maintenance Fees and Payment Method
3.1 Annual Maintenance Fee: USD 12,000 per hotel/resort per year. Party A operates a total of 20 hotels/resorts, so the total annual maintenance fee is USD 240,000 (excluding taxes). This fee includes SaaS multi-tenant maintenance of the NEXT SMS System, AI model retraining, Edge AI Gateway support, and training services.
3.2 Payment Schedule: A one-time annual transfer to the designated account of Party B.
3.3 Payment Method: Same as the Main Agreement (bank transfer to the designated account of Party B).
4. Rights and Obligations
4.1 Rights and Obligations of Party A:
| ● | Provide access rights to the NEXT SMS System (AI+SaaS), Edge AI Gateways, and related devices for each hotel/resort to facilitate maintenance; |
| ● | Provide anonymized video data for AI model retraining of the NEXT SMS System (if needed), ensuring compliance with PDPO; |
| ● | Make maintenance fee payments in accordance with Article 3; if the payment is overdue for more than 15 days, a penalty of 0.05% of the outstanding amount shall be charged daily. |
4.2 Rights and Obligations of Party B:
| ● | Provide maintenance services in accordance with Article 1, ensuring that the accuracy of the AI model of the NEXT SMS System remains ≥95% and the SaaS multi-tenant availability remains ≥99.9% during the maintenance period; |
| ● | Submit a monthly maintenance report to Party A, including the AI model performance dashboard (precision, recall rate) of the NEXT SMS System in each hotel and the SaaS multi-tenant fault statistics; |
| ● | Notify Party A in writing of potential risks (e.g., obsolescence of Edge AI Gateway hardware, compatibility between the SaaS module of the NEXT SMS System and new operating systems) 30 days in advance, and propose solutions (e.g., hardware replacement suggestions, software upgrade plans). |
17 / 18
5. Termination
In the event of a material breach by either party (e.g., Party A failing to pay maintenance fees for over 30 consecutive days, or Party B failing to resolve the NEXT SMS System AI model malfunction within 12 hours), the non-breaching party may issue a written notice demanding rectification. If the breaching party fails to correct the issue within 30 days, the non-breaching party may terminate this agreement and claim compensation for losses (e.g., damages resulting from prolonged downtime of the NEXT SMS System AI services).
6. Management law and dispute resolution
Same as the Main Agreement (subject to the laws of Hong Kong, disputes shall be settled through arbitration by the Hong Kong International Arbitration Center, which covers NEXT SMS system AI model maintenance breach compensation and SaaS multi-tenant service quality dispute issues).
Party A (Service Recipient): ALOHA Asia Pacific Limited
Authorized Signature: ____________________
Party B (Service Provider): Next Technology Holding Inc.
Authorized Signature: ____________________
18 / 18