Saad Alkentar

Developed an open-source, mobile-first ecosystem designed to provide an empathetic, therapeutic experience through AI-driven conversation. This project focuses on emotional regulation and accessibility, specifically targeting mental health support and individuals on the Autism Spectrum (ASD).

• Generative AI Integration: Integrated the Gemini API to simulate a therapist-like conversational style, enabling the diary to act as an empathetic listener rather than a passive log.
• Multimodal Accessibility: Engineered the mobile application (Flutter) with Speech-to-Text (STT) and Text-to-Speech (TTS) capabilities, allowing users to interact naturally through voice, which is critical for users with verbal communication challenges.
• Memory Capture & Recall: Developed a specialized 'Memory Capture' feature where the AI proactively prompts users for specific event details, later facilitating emotional 'reliving' and reflection through targeted reminders.
• Backend Architecture: Architected a robust Django backend to manage secure user accounts, diary persistence, and the orchestration of AI-generated responses.

Developed and deployed a centralized time and attendance platform to streamline workforce management. By integrating hardware with a custom web interface, the project reduced manual data entry errors by 36% and ensured real-time synchronization of employee records.

• Key Achievement: Engineered a device management extension to biometrically sync data from ZKTeco hardware (UA300 & IFace880) directly to a SQL Server backend.
• Scalability: Optimized database queries to handle high-frequency logs from multiple branch locations.
• Operational Impact: Reduced manual data entry errors by 36% through automated device synchronization and real-time validation.

BBSF Mobile is the flagship digital banking solution for Bank Bemo Saudi Fransi, engineered to provide customers with a secure, seamless, and comprehensive 'branch-in-your-pocket' experience. The application consolidates complex financial services into a user-friendly interface, enabling 24/7 banking operations ranging from basic account inquiries to advanced asset management.

• Comprehensive Banking Suite: Consolidated multi-tier financial services into an intuitive, secure interface enabling round-the-clock operations.
• End-to-End Asset Management: Facilitated full transaction lifecycles from instant fund transfers to complex corporate asset tracking.
• Enterprise Security: Implemented bank-grade encryption, secure token authentication, and multi-factor biometric verification.

Architected and deployed a comprehensive educational ecosystem that enables educators to monetize curriculum through digital assessments and interactive learning materials.

• Cross-Platform Delivery: Developed a high-performance student-facing mobile application using Flutter, optimized for seamless test-taking experiences and offline material access.
• Administrative Suite: Built a robust management dashboard using Vue.js for teachers to architect complex test structures, manage student enrollments, and track sales analytics.
• Scalable Backend: Engineered a centralized Django REST API to handle the business logic of content distribution, secure payment processing, and real-time result grading.
• Monetization Logic: Implemented a secure 'Test Bundle' marketplace, allowing for the creation, pricing, and automated delivery of educational materials to targeted student groups.

CPI is an AI-driven healthcare automation platform designed to streamline and accelerate prior authorization requests for specific medications. Built specifically for non-medical staff and administrative teams, the system reduces the complex, time-consuming burden of healthcare paperwork. By integrating directly with clinic FHIR (Fast Healthcare Interoperability Resources) systems, CPI extracts unstructured patient health records and uses artificial intelligence to structure the data, automatically populating the required medicine forms.

• FHIR Interoperability: Seamlessly connects with clinical electronic health records (EHR) to securely pull unstructured patient data, such as clinical notes, lab results, and diagnoses.
• AI-Powered Data Structuring: Utilizes Natural Language Processing (NLP) and Large Language Models (LLMs) to identify the specific medical criteria needed by insurance companies.
• Automated Form Population: Automatically fills out prior authorization medicine forms, minimizing manual data entry and reducing turnaround times for non-medical staff.
• Streamlined UI/UX: An intuitive, non-technical dashboard that allows administrative staff to review and submit paperwork without needing clinical expertise.

Developed an end-to-end healthcare ecosystem designed to digitize clinic operations while prioritizing patient data privacy and advanced searchability.

• Advanced Search & AI: Integrated Vector Embeddings to power a high-performance semantic search engine, allowing doctors to retrieve patient records accurately even with varied naming conventions or typos.
• Data Security & Privacy: Implemented AES-256 encryption for sensitive PII (Personally Identifiable Information) including patient names and contact details, ensuring data-at-rest security and healthcare standard compliance.
• Medical Imaging: Developed an image processing pipeline to handle and optimize medical imagery, enabling seamless viewing and management within the clinical dashboard.
• Full-Stack Architecture: Built a modular frontend with Nuxt.js (Vue.js) for a fast, SEO-friendly interface and a robust Django (Python) REST API backed by PostgreSQL for complex relational data management.

Dermatai is an advanced computer vision and medical image processing system designed to assist dermatologists and healthcare providers in the accurate detection and monitoring of skin lesions. By leveraging deep learning and image analysis techniques, the system provides an automated pipeline to segment lesions, calculate their precise dimensions, and track their progression over time.

• Automated Lesion Segmentation: Utilizes the U-Net architecture, a convolutional neural network optimized for biomedical image segmentation, to accurately identify and isolate lesion boundaries from patient photos.
• Accurate Size Estimation: Integrates a reference sticker detection algorithm to calibrate the image and estimate the real-world dimensions of the lesion, ensuring clinical-grade accuracy regardless of camera distance or angle.
• Longitudinal Progress Tracking: Automatically plots and visualizes lesion progression over time based on consecutive patient-uploaded images, allowing dermatologists to monitor growth or regression across specific dates.

Developed a comprehensive event planning ecosystem designed to streamline the workflow for organizers and end-users. Eventac simplifies complex planning by breaking events into granular, manageable actions.

• Project Management Architecture: Engineered a hierarchical task system allowing users to decompose events into specific tasks and sub-steps for precision tracking.
• Vendor Integration: Built a seamless marketplace interface where users can discover service providers, book appointments, and link specific vendor products directly to project steps.
• End-to-End Orchestration: Integrated real-time scheduling and resource allocation features to ensure a seamless transition from the 'Idea' phase to 'Execution.'

Developed a sophisticated biometric identification system combining facial recognition and fingerprint analysis, achieving a validated 98% accuracy rate across a benchmark dataset of 4,320 images.

• Facial Embedding Engine: Engineered a pipeline to extract and embed high-dimensional facial features into a vector space, enabling rapid and precise identification.
• Mathematical Optimization: Implemented advanced vector distance functions (Cosine Similarity/Euclidean Distance) to perform nearest-neighbor searches, ensuring high-speed retrieval from the database.
• Multi-Modal Integration: Developed a secondary authentication layer using fingerprint recognition, increasing system security and reducing False Acceptance Rates (FAR).
• Data Validation: Conducted extensive testing and performance tuning on a large-scale dataset, optimizing the threshold for feature matching to hit peak precision and recall metrics.

Developed a high-performance e-commerce platform dedicated to the 'Honey Spoon' product line, focusing on a seamless user journey from discovery to checkout. Engineered a robust backend architecture to handle secure transactions and real-time inventory management.

• Transactional Pipeline: Engineered a robust backend architecture to handle high-concurrency transactions, automated order routing, and real-time stock reconciliation.
• Strategic Payment Pivot: Successfully led a strategic pivot to prioritize PayTabs integration for the MVP, ensuring a stable and secure launch while maintaining a modular codebase ready for the upcoming PayPal rollout in Version 2.0.
• Modular Service Architecture: Decoupled catalog management from financial pipelines to enable swift feature iteration and promotional customization.

I am The Chef is a comprehensive digital ecosystem designed to bridge the gap between home cooking and professional dining. This full-stack application serves as a central hub for culinary enthusiasts to discover, share, and explore the world of food.

• Digital Ecosystem: Bridged home cooking and culinary gastronomy through interactive recipe publishing, cooking timers, and dietary filters.
• Full-Stack Cross-Platform: Built responsive mobile applications for Android and iOS coupled with high-speed RESTful APIs for media sharing and community interaction.
• Social Discovery Engine: Implemented recipe bookmarking, chef following, and interactive search algorithms for culinary inspiration.

Designed and developed a high-performance management interface and backend for the I-Flow platform, enabling the integration of IT and OT (Operational Technology) systems within a Unified Namespace (UNS).

• Backend Architecture: Engineered a modular, enterprise-grade backend using NestJS (Node.js), leveraging its dependency injection and microservices capabilities to manage complex industrial data flows.
• Frontend Interface: Built a sleek, responsive control plane using Vue.js (Nuxt.js), focusing on real-time data visualization and intuitive configuration for factory-scale deployments.
• Protocol Simulation: Developed a specialized data generation engine to simulate MQTT and AMQP traffic, allowing for the stress-testing of industrial brokers and validation of data harmonization logic without requiring physical hardware.
• Real-Time Messaging: Implemented pub/sub patterns to handle high-frequency data operations, ensuring low-latency communication between simulated devices and the central hub.

Engineered a high-performance portal designed to bridge web-based management with a mobile app. Leveraged a modern asynchronous backend and a reactive frontend to ensure real-time data flow and system responsiveness.

• Frontend Excellence: Built a dynamic, responsive user interface using React, focusing on state management and seamless user workflows for device monitoring.
• Backend Performance: Developed a robust, asynchronous API layer using FastAPI, optimizing data throughput and ensuring low-latency communication between the portal and connected iOS endpoints.
• DevOps & Orchestration: Containerized the entire microservices architecture using Docker and managed deployment via Kubernetes, ensuring high availability, automated scaling, and environment consistency.

Architected a cloud-native backend on AWS to automate saltwater aquarium management and water chemistry tracking. The system utilizes Optical Character Recognition (OCR) to digitize physical test reports into a chronological health database.

• Serverless Architecture: Designed a scalable, event-driven backend using AWS Lambda and API Gateway, ensuring zero-cost idling and high performance during bursts.
• Data Pipeline & Automation: Integrated SQS (Simple Queue Service) and SNS (Simple Notification Service) to decouple report processing tasks, ensuring reliable data delivery and real-time user alerts.
• AI/OCR Integration: Implemented Tesseract OCR within a Lambda environment to automatically extract chemical parameters (pH, Salinity, Alkalinity) from uploaded report images stored in Amazon S3.
• Security & Persistence: Managed user authentication and secure document storage, maintaining a historical time-series record of aquarium health for long-term trend analysis.

Developed a cutting-edge fitness application that utilizes Computer Vision to provide real-time form correction and performance feedback for gym-goers.

• Computer Vision Integration: Leveraged MediaPipe for high-fidelity pose estimation, tracking body keypoints to analyze exercise form in real-time.
• Pattern Recognition: Engineered custom algorithms to recognize movement patterns, enabling the system to count repetitions and identify technical errors in user performance.
• Cross-Platform Development: Architected a fluid, high-performance mobile interface using Flutter, ensuring a consistent experience across iOS and Android.
• Direct Feedback Loop: Designed the logic for 'Direct Feedback,' transforming raw pose data into actionable insights for users to prevent injury and optimize training results.

Developed high-reliability Android and iOS applications designed to help parents track infant feeding patterns and nursing data with 100% data persistence.

• Mobile Development: Engineered the native Android application from the ground up using Kotlin, focusing on a seamless and empathetic user experience.
• Offline-First Architecture: Implemented a robust local-storage strategy using SQLite, ensuring users could log critical data instantly regardless of internet connectivity.
• Data Synchronization: Designed a background sync engine to automatically reconcile local SQLite data with the remote backend API once a connection is restored, resolving potential data conflicts.
• Performance: Optimized local database queries to ensure zero-latency logging, which is critical for real-time tracking in high-stress parenting scenarios.

Engineered a comprehensive job board ecosystem designed to bridge the gap between employers and talent. Built a scalable architecture to handle complex subscription models (bundles) and automated job-matching logic.

• Employer Suite: Architected a monetization system using 'Vacancy and Resume Bundles,' allowing companies to manage recruitment budgets while accessing a filtered talent pool.
• Job Seeker Suite: Developed a dynamic resume builder and an automated 'Job Agent' notification engine to alert users of high-relevance opportunities based on specific criteria.
• Core Functionality: Implemented advanced matching algorithms to correlate candidate qualifications with active job requirements, significantly reducing time-to-hire.

Designed and implemented a three-tier synchronization ecosystem for ZKTeco biometric hardware, pivoting from a 'Pull' to a 'Push' data architecture to improve real-time data reliability.

• Edge Software: Developed a custom scheduler resident on ZKTeco hardware (UA300 & IFace880) to autonomously initiate data transfers to a central server.
• Deployment Engine: Created a specialized installation utility to streamline the deployment of the scheduler across multiple network-connected devices.
• Backend & API: Engineered a high-concurrency web service to receive, validate, and commit incoming device data to a SQL Server database, ensuring data integrity across large-scale deployments.