
Closed
Posted
Paid on delivery
Project Title: AI-Based Smart Shoplifting Detection System Role: AI & Full-Stack Developer (Remote Implementation) 1. Project Overview We are building a smart surveillance system that uses Edge AI to detect shoplifting in real-time. The system runs locally on an NVIDIA Jetson Orin Nano, processes video feeds from an NVR/existing cameras, and upon detecting a theft, instantly sends a 20-second video clip to a mobile app. We require a developer to build the software stack, implement the system remotely, and optimize the AI models to achieve a minimum of 85% detection accuracy in addition to be teachable and self learnable and improvable. 2. System Architecture & Components • Hardware (Provided by Client): o NVIDIA Jetson Orin Nano (Edge Processing). o NVR with RTSP streams. o Existing IP/CCTV Cameras (Various brands). • Software (To be developed by You): o Edge AI Module: Computer vision model running on Jetson Orin Nano. o Mobile App: For receiving alerts and viewing clips. o Web Dashboard: For managing multiple systems/locations. 3. Detailed Scope of Work (Deliverables) A. Edge AI Development (Jetson Orin Nano) • Develop and deploy a computer vision model optimized for the Jetson Orin Nano . • Functionality: o Ingest RTSP streams from the NVR/Cameras. o Detect shoplifting behaviors (e.g., concealing items, irregular movements). o Clip Generation: Automatically cut a 20-second video clip (approx. 5s before event + 15s after) the clip size should be compresses in order to quick transmitting. o Select the needed cameras: not all the cameras in the network should be in the system , the system should support selecting only needed Ips or the number of camera. o The system must not be static; it must improve based on the Manager's feedback from the Mobile App but this improving should be only locally and doesn’t affect in other branches or brnads. This is due to different policies in every market (some branches account hiding in the private bag as a theft while others not. So the feedback improving should be only locally. • Security: The source code deployed on the Jetson must be fully encrypted/obfuscated to prevent reverse engineering. • Quick and teachable : The clip should be sent to the mobile app within 10s or less with tag right alert/false alert to accept the feedback and improve the response. • Controlled : Runing the system should be controlled by the main dashboard in addition to archiving all the history in that dashboard. B. Mobile Application (iOS & Android) • Develop a secure mobile app for the multi–Store Manager. • Features: o Receive real-time push notifications of theft incidents. o Play the specific 20-second evidence clip received from the local system. o Notifications should be sent within max 10s and it must include two separate buttons with different colors right alert (green) / false alert (red). o Running the app should be according to right user and password and should be activated or deactivated from the main dashboard. o Alert the admin if a camera stream is lost or show the connected to system cameras status. C. Central Web Dashboard • Develop a web-based main dashboard to observe multiple systems. • Features: o Status monitoring (Online/Offline) of the Jetson devices. o Incident logs according to the branch and name of the market o The system should support unlimited branches , the system should be adding according to the following tree and data (brand name => branch name => location address) o Managers feedback to monitoring the efficiency of system in every branch as a gaph and numeric data o User Access: A table of App Users with "Enable/Disable" toggles. Revoking access stops alerts to that phone immediately. o • System Toggle: A switch to Remotely Activate/Deactivate the AI processing on the Jetson (e.g., during store closing hours). o • support creating main dashboard users with selected branch monitoring . D. Remote Implementation & Configuration • Since this is a remote contract, you are responsible for: o Providing a guide and training for the initial hardware setup. o Remotely accessing the Jetson device (via SSH, VPN, or remote desktop tools) to install dependencies, deploy the code, and configure the camera streams. o Troubleshooting connection issues between the NVR/cameras and the AI module remotely. 4. Performance & Acceptance Criteria The project will only be considered complete upon meeting the following criteria: 1. Accuracy Guarantee: The system must achieve a minimum firstly 85% accuracy rate in detecting shoplifting events during the testing phase. You are expected to fine-tune the model thresholds to balance False Positives and False Negatives. 2. Latency: The alert and video clip must be received on the mobile app within reasonable latency (near real-time) after the event occurs. 3. Security: Verification that the deployed code is encrypted and cannot be easily accessed or copied from the hardware. Summary of Acceptance Criteria for the Developer 1. Accuracy: >85% (validated by Manager Feedback). 2. Speed: <10 seconds from event capture to App Alert. 3. Control: Dashboard must successfully Pause/Resume the remote hardware. 4. Feedback: App must successfully send "Correct/Wrong" tags to the database. 5. Security & IP Protection (Strict) To prevent theft of the code and AI models, the following protections are required: Code Compilation: No raw Python (.py) files. All code must be compiled (e.g., Cython to .so binaries). Disk Encryption: The Jetson's storage must use LUKS Full Disk Encryption. Hardware Locking: The software must verify the device's Unique Hardware ID/Serial Number at startup. If copied to another board, it must fail to run. 6. Milestones & Payment Schedule Phase 1: Mobilization & System Architecture (20%) • Deliverables: o Final Technical Design Document (System Architecture & Data Flow). o Setup of the Cloud Infrastructure (Server/Database). o Establishment of Remote Connection to the Jetson Hardware. o Successful ingestion of RTSP streams from NVR/Cameras into the Jetson. • Payment Trigger: demonstration of camera feed access on the Jetson. Phase 2: MVP Development (Core Pipeline) (20%) • Deliverables: o Edge AI: Basic model deployment detecting "Person" and "Motion" on the Jetson. o Mobile App: Beta version released (iOS/Android) capable of receiving a push notification. o Pipeline Test: Proof that a triggered event sends a video clip to the phone (regardless of accuracy or speed at this stage). • Payment Trigger: Video evidence of a "Detection -> Notification -> Video Playback" loop. Phase 3: Advanced Features & Security Implementation (30%) • Deliverables: o Latency Optimization: Achieving the <10-second alert speed. o Dashboard Control: Implementation of the "Kill Switch" and User Management. o Self-Teaching: "True/False" feedback buttons functional and logging to the database. o Security: Implementation of Code Compilation (Cython) and Disk Encryption. • Payment Trigger: Demonstration of the "Kill Switch" working instantly and verification of encrypted code. Phase 4: Optimization, Accuracy & Final Handover (30%) • Deliverables: o Accuracy Tuning: Fine-tuning the model to achieve >85% accuracy on live store data. o Stress Testing: System stability test (running 48 hours without crash). o OTA Update Test: Successful remote push of a new model version. o Documentation: User Manuals and Admin Guides. • Payment Trigger: Final Acceptance Sign-off confirming 85% accuracy and full system stability for 48 hours. Notes : 1- It will be a specialized third party auditing the deliverable quality. 2- There is an attached file has more information about the project.
Project ID: 40181101
112 proposals
Remote project
Active 2 mos ago
Set your budget and timeframe
Get paid for your work
Outline your proposal
It's free to sign up and bid on jobs
112 freelancers are bidding on average $444 USD for this job

Hello, I will build an edge-first shoplifting detection stack for the Jetson Orin Nano that ingests RTSP streams from your NVR/cameras, runs a compact, fast detection model, and creates a 20-second evidence clip (about 5 seconds pre-event, 15 seconds post-event) to send to a mobile app within 10 seconds. The system will allow selecting only needed cameras, support remote deployment, and learn locally from manager feedback so policies stay branch-specific. To protect your IP, the code will be compiled (e.g., via Cython) and the storage encrypted; the dashboard will remotely control processing, archive history, manage users, and monitor connections. I will deliver a secure, modular solution with a clear API for the web dashboard, iOS/Android app, and remote setup, plus thorough documentation and training. 1) What is the target accuracy and how will you validate it across stores? 2) How many cameras/branches will be onboarded and can you provide a camera list? 3) What privacy constraints apply to video clips and logs, and who can access them? 4) Do you need OTA updates, version control, and rollback capability for the edge module? 5) What security standards must be met for on-device encryption and anti-tamper checks? Best regards,
$550 USD in 27 days
9.0
9.0

Hello, I am an AI and full stack developer with 8 plus years experience building edge AI systems and secure real time pipelines. I can deliver your smart shoplifting detection system on NVIDIA Jetson Orin Nano with mobile alerts and a central dashboard, meeting your accuracy, latency, and security requirements. Wordflow in points First I will design the full system architecture and set up cloud infrastructure, remote access to Jetson, and RTSP ingestion from NVR cameras. Next I will develop the edge AI pipeline, deploy an optimized model on Jetson, and generate 20 second clips with pre and post event buffering. Then I will build the mobile app for iOS and Android with real time push notifications, clip playback, and correct false alert feedback buttons. After that I will create the central web dashboard for multi branch monitoring, user access control, kill switch activation, and incident logs with analytics. I will implement local learning so feedback improves the model only per branch without affecting other locations. Security will be enforced through compiled code, LUKS disk encryption, and hardware binding to prevent copying. Finally I will optimize latency to under 10 seconds, tune the model to achieve over 85 percent accuracy, and provide full documentation, OTA update support, and a 48 hour stability test for final acceptance. Thanks Julian
$250 USD in 7 days
8.4
8.4

Hello hope you doing well i will help you build a real time AI based smart shoplifting detection system using NVIDIA Jetson Orin Nano that reliably detects theft events with over 85% accuracy and delivers 20 second evidence clips to your mobile app within 10 seconds. i will develop an edge AI module that ingests RTSP streams, supports selective cameras, and learns locally from manager feedback while keeping code fully encrypted and hardware locked. the mobile app will securely notify managers with correct/false alert buttons, and the central web dashboard will monitor multiple branches, manage users, toggle AI processing, and log incidents for analysis. you will receive a fully optimized secure and scalable system with remote deployment support technical documentation and a seamless workflow from detection to alert and feedback integration. Discover my professional portfolio: https://www.freelancer.com/u/Hammadhassan21 If my offer meets your expectations, we would be happy to discuss your project in more detail. Best Regards, Hammad Hassan
$400 USD in 7 days
7.3
7.3

Hello. I propose to design and implement a secure, edge-based AI shoplifting detection platform optimized for NVIDIA Jetson Orin Nano, integrating real-time RTSP ingestion, low-latency event detection, and automated evidence clip generation. The Edge AI pipeline will leverage optimized deep-learning models (TensorRT-accelerated) with behavior-level inference and adaptive thresholds, enabling >85% accuracy while supporting localized, branch-specific learning via manager feedback. A secure mobile application (iOS & Android) will deliver near real-time alerts (<10s) with encrypted clip playback and explicit True/False feedback mechanisms to continuously improve local model performance. A centralized web dashboard will provide multi-brand, multi-branch system control, real-time device status, incident analytics, and fine-grained user access management. All deployed software will be hardened using compiled binaries (Cython), hardware-bound license checks, and LUKS full-disk encryption to protect IP and prevent unauthorized duplication. The system will support remote provisioning, OTA model updates, and fail-safe device control, including instant AI activation/deactivation from the dashboard. Performance, security, and accuracy targets will be validated through live testing, stress runs, and third-party audit compliance. The final delivery will include complete documentation, operational training, and a production-ready system meeting all acceptance criteria. Regards, justin.
$400 USD in 7 days
6.6
6.6

As CodeNomad, we're a seasoned team of AI and Full-Stack developers with 12+ years' experience in delivering cutting-edge software solutions like your AI Shoplifting Detection System. Our proficiency with Python, which is crucial for the development and optimization of your Edge AI Module for the Jetson Orin Nano, makes us an excellent fit for this project. We're well-versed in working with NVIDIA devices and can ensure optimal performance to achieve the minimum detection accuracy of 85% you require. With our skills in Mobile Development, we'll design and develop a secure mobile application that aligns with your needs to provide real-time push notifications, play specific video clips, and give Managers fast response options ensuring efficient store management. Similarly, our expertise in Web Development will let us craft a well-designed and user-friendly main dashboard that not only observes multiple systems but also gives you centralized control over all branches. One aspect where we shine is deep learning model deployment which is essential for the advanced shoplifting behavior detection algorithms enabling your system to get smarter over time. As you've explicitly noted about the importance of security in code deployed on Jetson Nano, rest assured, our team is highly attentive to encryption and obfuscation techniques ensuring utmost security. Thanks.....
$500 USD in 7 days
6.3
6.3

Hello There!!! ⚜⭐⭐⭐⭐⚜(( Build a secure Edge AI shoplifting detection system with real time alerts ))⚜⭐⭐⭐⭐⚜ Your project is a full Edge AI pipeline on Jetson Orin Nano, streaming from NVR cameras, detecting shoplifting behaviors, and sending 20 second evidence clips to a mobile app. The system must be mobile responsive, accurate, and secured with code encryption and hardware locking. I have strong experience building real time video pipelines, deploying AI models on edge devices, and developing secure web dashboards with mobile apps. I will structure the system in clear milestones, starting with RTSP ingestion, then detection, clip generation, and finally security and feedback learning. My mini strategy is to use optimized YOLO or EfficientDet on Jetson, implement local feedback learning, compress clips, and connect the dashboard for remote control and monitoring. Top 3 features: 1. Mobile alerts with evidence clip in under 10 seconds 2. Secure compiled code with hardware binding 3. Dashboard control with feedback and analytics If you share the attached file and access details, I can start the architecture and milestone plan immediately. Warm Regards, Farhin B.
$256 USD in 10 days
6.6
6.6

Hello, I trust you're doing well. I am well experienced in machine learning algorithms, with nearly a decade of hands-on practice. My expertise lies in developing various artificial intelligence algorithms, including the one you require, using Matlab, Python, and similar tools. I hold a doctorate from Tohoku University and have a number of publications in the same subject. My portfolio, which showcases my past work, is available for your review. Your project piqued my interest, and I would be delighted to be part of it. Let's connect to discuss in detail. Warm regards. please check my portfolio link: https://www.freelancer.com/u/sajjadtaghvaeifr
$1,800 USD in 7 days
6.4
6.4

Hello! As per your project post, you are looking to build an AI-based Smart Shoplifting Detection System that runs on NVIDIA Jetson Orin Nano devices, processes live video from selected cameras, and instantly delivers 20-second evidence clips to a mobile app. The goal is to provide branch-specific, locally improving AI detection while offering a secure, centralized web dashboard for monitoring, feedback, and system control. My focus will be on delivering a reliable Edge AI solution where each store can independently detect shoplifting behaviors with fast, accurate alerts. The Jetson module will process only selected camera streams, generate compressed clips for near real-time delivery, and incorporate local feedback from managers to improve detection without affecting other branches. The mobile app will handle secure notifications, quick playback of clips, and structured feedback (right/false alert). I specialize in building operational AI systems that integrate edge devices, real-time notifications, and web dashboards. My approach emphasizes low-latency performance, local adaptability, and strong encryption to protect intellectual property while keeping the system modular and scalable across multiple branches. Let’s connect to align on deployment priorities, feedback handling, and system scalability so we can deliver a secure, high-performing shoplifting detection system tailored for your multi-branch operations. Best regards, Nikita Gupta
$500 USD in 45 days
6.1
6.1

Hi there,Good morning I am Talha. I have read you project details i saw you need help with Mobile App Development, Deep Learning, Video Processing, Full Stack Development, Computer Vision, Web Development, AI Development and Security I am writing to propose an innovative approach to tackle your project. Our proposal centers on delivering creative and effective solutions that will set your project apart. We will present fresh, out-of-the-box ideas that align with your project's objectives, demonstrating how we can achieve remarkable results. Please note that the initial bid is an estimate, and the final quote will be provided after a thorough discussion of the project requirements or upon reviewing any detailed documentation you can share. Could you please share any available detailed documentation? I'm also open to further discussions to explore specific aspects of the project. Thanks Regards. Talha Ramzan
$250 USD in 10 days
5.6
5.6

Hi, I've thoroughly reviewed your project requirements for the AI-Based Smart Shoplifting Detection System and am confident in delivering a robust solution. I have practical experience working with LLMs, Transformers, and real-time computer vision deployment on edge devices like NVIDIA Jetson. My background includes building AI-driven video processing pipelines, integrating secure mobile apps, and developing scalable web dashboards. I understand the complexity of balancing detection accuracy, latency, and user feedback to create a self-learning system tailored for each store while maintaining strong code security with compilation and encryption. I propose starting with a detailed technical plan and remote setup guidance followed by iterative development and testing phases as outlined, ensuring the system meets your 85% detection accuracy and under 10-second alert requirements. I am ready to collaborate to refine feedback loops and optimize performance upon deployment. Looking forward to discussing the specific camera models you plan to use and your preferred technologies for the dashboard and mobile app. Could you specify which mobile app frameworks or technologies you prefer for this project? Thanks, Andrew
$555 USD in 5 days
5.7
5.7

As a seasoned Full-Stack Developer with deep expertise in Mobile App Development, I am thrilled about the opportunity to apply my skills in creating your AI-based smart shoplifting detection system. I possess significant expertise in Computer Vision and Machine Learning, skills that are fundamental for developing an efficient edge AI solution, capable of accurately detecting shoplifting behaviors in real-time and generating 20-second video clips. In addition to my technical proficiency, I bring with me over six years of remote working experience; adept at troubleshooting, implementing robust solutions while catering to different time zone preferences. Furthermore, I 'm well-versed in enhancing systems post-deployment. It is critical the system can adjust to local feedback without impacting other branches or markets; this request aligns perfectly with my approach towards dynamic solutions development. My capacity to architect reliable and scalable applications will be instrumental in building a robust web-dashboard that manages multiple systems/locations efficiently, providing centralized monitoring while maintaining uncompromising data security. Additionally, I have solid experience in AWS DevOps, ensuring efficient deployment and management of cloud resources for optimised system performance.
$400 USD in 7 days
5.9
5.9

Hello, I am excited about the opportunity to collaborate on developing your AI-based smart shoplifting detection system. As a top California freelancer with numerous five-star reviews, I specialize in AI and full-stack development, and I'm confident I can deliver a robust solution tailored to your unique needs. I understand the critical importance of achieving at least 85% detection accuracy while ensuring the system can learn and improve locally based on feedback from store managers. I will leverage my expertise to create a seamless Edge AI module that efficiently ingests RTSP streams, detects shoplifting behaviors, and generates quick video alerts to the mobile app. The web dashboard will facilitate comprehensive control over the entire system, ensuring secure monitoring across multiple branches. Let’s discuss your project further so we can ensure all your specifications are met efficiently. What specific features do you envision for the user feedback functionality in the mobile app? Best regards,
$490 USD in 10 days
5.1
5.1

Hello client, I'm Denis Redzepovic, an experienced developer with expertise in Security, Full Stack Development, Deep Learning, AI Development, Video Processing, Web Development, Mobile App Development and Computer Vision. I have worked extensively on diverse Python projects, ranging from backend development and automation to data processing and API integrations. My deep understanding of Python’s libraries and frameworks allows me to build efficient, scalable, and maintainable solutions. I pay close attention to code quality and performance to ensure your project runs flawlessly. With my solid experience, I’m confident I can deliver results that exceed your expectations. I focus on writing clean, maintainable, and scalable code because I know the difference between 99% and 100%. If you hire me, I’ll do my best until you’re completely satisfied with the result. Let’s discuss your project details so I can tailor the perfect Python solution for you. Thanks, Denis
$350 USD in 7 days
5.0
5.0

⚠️You are not looking for a coder. You are looking for someone who can build this properly. That is exactly why your project stood out.⚠️ Your vision for an AI-based smart shoplifting detection system leveraging Jetson Orin Nano and a multi-layered architecture reflects a commitment to precise, scalable edge intelligence that adapts locally without compromising security. That aligns directly with how we engineer systems at DigitaSyndicate — developing intuitive, secure, and future-proof automated platforms. At DigitaSyndicate, a UK-based digital systems agency, we build precision-engineered automation, modern web platforms, and AI-driven systems designed for performance and long-term scalability. Our approach to encryption, hardware locking, and distributed feedback loops ensures the robustness and controlled adaptability your system requires. We recently delivered a real-time video analytics platform with sub-10 second latency and encrypted edge deployment. Can you share your main priorities and timeline so I can map out the right execution plan for you? Casper M. Project Lead | DigitaSyndicate Precision-Built Digital Systems.
$400 USD in 14 days
5.0
5.0

Hi There!!! The Goal of the project:- Develop an AI-based smart shoplifting detection system with Edge AI on NVIDIA Jetson Orin Nano, a mobile alert app, and a central web dashboard for real-time monitoring and feedback. I have carefully read your detailed project description and understand the need for a secure, low-latency, self-improving shoplifting detection system that integrates with multiple cameras, provides mobile alerts, and central dashboard control. I am the best fit for this project because I combine expertise in AI/Computer Vision, full-stack development, and secure embedded systems deployment to deliver reliable end-to-end solutions. 1. Edge AI module development on Jetson Orin Nano with local video processing and clip generation 2. Mobile application for iOS and Android with real-time alerts and feedback buttons 3. Central web dashboard for monitoring, user access control, and remote activation/deactivation of the AI system I provide UI design, database management, testing, and full source code delivery at project completion. I have 9+ years experience as a full stack developer and have completed similar AI surveillance and real-time video processing projects with mobile and dashboard integration. Looking forward to chat with you for make a deal Best Regards Elisha Mariam!
$269 USD in 6 days
5.0
5.0

Hello! I'm excited about your project on developing an AI-Based Smart Shoplifting Detection System. I understand you need a robust solution that leverages Edge AI, specifically on NVIDIA Jetson Orin Nano, to detect theft in real-time and send alerts to a mobile app while ensuring high accuracy and security. My approach will focus on developing an optimized computer vision model that meets your detection criteria. I will implement a user-friendly mobile application enabling immediate notifications and oversight through a central web dashboard, all while ensuring data security and compliance with your specifications. Regards, Davide
$450 USD in 7 days
4.8
4.8

Hello, I’ve read your project “AI Shoplifting Detection Developer Needed” and understand what you’re aiming to achieve. I can deliver a clean, reliable result with clear communication and quick turnaround. https://www.freelancer.com/u/proggon Best regards, Wahaj Barlas
$400 USD in 7 days
4.7
4.7

Hi, I'm excited about the opportunity to develop your AI-Based Smart Shoplifting Detection System. With over 10 years of experience in AI and full-stack development, I am well-equipped to build a robust system that meets your needs. I will design an optimized Edge AI module to effectively process RTSP video feeds on the Jetson Orin Nano, ensuring we achieve the required detection accuracy. My strong proficiency in computer vision and mobile app development will ensure real-time alerts and seamless integration. I plan to collaborate closely with you to tune models for optimal performance and security, delivering on-time and with quality. Best regards, Volodymyr
$500 USD in 2 days
4.7
4.7

As an experienced Machine Learning Engineer with a strong focus on computer vision, I can confidently say that I have the necessary skills and expertise required to develop your AI-based Smart Shoplifting Detection System. Over my 8+ years in the industry, I have designed clean data pipelines, optimized model accuracy and speed, and implemented scalable AI systems that effectively recognized complex patterns. More specifically, my specialization in deep learning is especially pertinent for this project as it aligns directly with the demands of training and optimizing an efficient computer vision model for your NVIDIA Jetson Orin Nano. In terms of frontend development, I am proficient in both iOS and Android app development and can deliver a secure, user-friendly mobile app for multi-store managers to receive real-time push notifications, access footage and provide invaluable feedback. In addition, my web development skills will come in handy as I create an easily navigable central dashboard where you can monitor multiple systems' statuses, access incident logs based on branch names, along with the efficiency analysis drawn from each branch's data graphically and numerically. With me on board, your edge AI module, mobile application, web dashboard, branch expansion plans and more will be in great hands. Let's get started!
$400 USD in 7 days
4.8
4.8

I'm experienced in developing AI-based systems and optimizing computer vision models for real-time surveillance. I have expertise in building Edge AI modules, mobile apps, and web dashboards. My focus is on achieving high detection accuracy, quick response time, and seamless integration of security features. I ensure the system is self-learning and customizable based on user feedback. Let's collaborate to create an efficient shoplifting detection solution.
$400 USD in 7 days
4.3
4.3

Jeddah, Saudi Arabia
Payment method verified
Member since Jan 14, 2026
$250-750 USD
$30-250 USD
$30-250 USD
$30-250 USD
$30-250 USD
$250-750 USD
₹12500-37500 INR
₹37500-75000 INR
₹600-7000 INR
$2-8 USD / hour
$3000-5000 AUD
₹600-1500 INR
₹12500-37500 INR
$1500-3000 USD
$3000-5000 USD
₹100-400 INR / hour
$3000-5000 USD
₹750-1250 INR / hour
₹150000-250000 INR
₹600-1500 INR
€30-250 EUR
$15-25 USD / hour
₹12500-37500 INR