Speech Therapy Platform
Audio AI
Pronunciation Recognition
A custom speech recognition system for children’s therapy, built with phoneme-level modeling, fine-tuned datasets, and production-ready deployment.
Cases
Doug Labs
On-Device Neurological Assessment for iOS
Key results
- Streamlined neurological test assessment in one intuitive app.
- Delivered accurate pronator drift, leg drift, and eye movement tests.
- Provided enhanced security via data encryption, Face ID authentication, screen capture protection.
- Ensured no cloud dependencies and complete user data privacy.
Services- Computer VisioniOS Development
Industry- Healthcare
TECHNOLOGIES USED- MediaPipeWhisperPythonSwift
Team- 1 AI Engineer1 iOS Developer1 AI Tech Lead1 Project Manager
Duration- 6 months / ongoing
Location- USA
The Challenges & Goals
Traditional neurological testing relies on manual observation and recording, which is time-consuming, subjective, and prone to human error. The client wanted to digitize and streamline this process with an AI-powered mobile app capable of performing neurological testing. This task demanded balancing clinical accuracy, patient privacy, and real-world usability within a single mobile solution, where our team had to solve the following challenges:
- 01
On-device neurological testing
Developing an AI system that performs movement assessments with local data processing and storage
- 02
Medical-grade accuracy
Achieving precise limb (arm, leg) and eye movement tracking and ensuring reliable performance under varying lighting and camera conditions.
- 03
Intuitive app workflow
Designing a user-friendly interface that guides both neurologists and patients through each test step with minimal training
- 04
Robust data security
Implementing app security, data encryption, and other similar measures to meet healthcare standards
- 05
Lay the groundwork for scale
Establishing a foundation for expanding the product with additional neurological tests and potential integrations with medical systems
Solution Overview
The It-Jim team developed a custom AI-driven iOS application that performs on-device neurological assessments using advanced computer vision, machine learning algorithms, and iOS optimization. The step-by-step approach ensured the final solution is accurate, privacy-compliant, and ready for real-world medical use. Doug Neuro AI app enables doctors to perform core neurological tests – pronator drift, leg drift, and eye movement – quicker, standardized, and accessible anytime, anywhere.
- Stage 01
Understanding Product Peculiarities & Research
In the discovery phase, our team collaborated closely with Doug Labs’ neurologist to translate diagnostic expertise into quantifiable tests and tasks that AI could automate.
We began our cooperation with in-depth consultations to d
Key initial steps by It-Jim:
- Collecting technical and functional requirements for the future app
- Researching testing procedures based on neurological standards
- Defining main neurological tests and their metrics (e.g., angle, symmetry, duration)
- Defining the main app flow and user experience
- Highlighting the requirements for personal data protection
With the completion of this first milestone, we have successfully established a clear roadmap for building a secure iOS app: product backlog, system architecture, and tech documentation. This milestone helped validate the concept and functionality of an AI-assisted mobile assessment tool for neurology.
- Stage 02
Feasibility Testing of On-Device AI Tracking
With a clear understanding of the project flow, we developed a working prototype to validate the AI pipeline for accurately detecting and tracking body and eye movements using a smartphone camera.
The app needed to run entirely on-device, so we used MediaPipe to track body landmarks and movements locally. After test recording, the app processes the MediaPipe data to analyze speed, direction, rotation, and symmetry of the patient’s arms, legs, and eyes. It then generates outcomes with a severity rating and saves the results in a JSON file.
The PoC demonstrated and achieved:
- Real-time landmark detection of arms, legs, and eyes using live video input
- Accurate movement analysis – speed, direction, and symmetry
- Reliable local operation without external servers or cloud computing
- Validated that all AI processing could run locally, on-device, maintaining privacy and responsiveness
This phase demonstrated that a lightweight mobile device can deliver clinically useful accuracy in a set of neurological tests.
- Stage 03
MVP Development & Launch
After validating the PoC, we developed and launched the MVP - a ready-to-pilot version for healthcare professionals. MVP included three integrated neurological tests:
Pronator drift test, Leg drift test, Eye movement test, Double Simultaneous Stimulation Speech test.
Following rigorous testing, we successfully launched the MVP. This milestone represents a significant advancement for neurologists, offering them a portable and secure tool for early patient assessment and evaluation.Core app features included
- Real-time AI analysis with automatic abnormality detection (normal/abnormal, severity, and affected side)
- Voice recognition
- Secure authentication (Face ID) and data encryption
- Screen capture protection for extra confidentiality
- Step-by-step test instructions and visual feedback for consistent results
- Test result summaries with severity detection and abnormality side
- Notifications and real-time alerts for improper framing or test setup
- Automatic real-time feedback and alerts
The MVP delivers instant in-app results from three main neurological tests, along with a structured JSON file containing diagnosis data, all processed locally for maximum patient security and privacy.
- Outcome
Final Outcome: On-Device Neurological Patient Assessment
Our cooperation with Doug Neuro AI has introduced a unique digital, data-driven product for neurological examination, helping clinicians evaluate motor and eye functions quickly, consistently, and securely.
Key achievements
- Fast MVP delivery in under 4 months
- Eliminate data privacy risks through on-edge AI
- Achieve accurate eye movement detection on iOS via a custom-built MediaPipe pipeline
- Integrate end-to-end app security with data encryption, Face ID authentication, and screen capture protection
- Simplify neurological assessments - practitioners can now perform testing within a single app
- Lay the foundation for clinical validation and government partnerships
- Establish a foundation for integrating with healthcare systems and expanding neurological tests
Result: This collaboration demonstrated how AI, computer vision, and on-device computing can accelerate neurological care, making it faster, safer, and more accessible. As a result, Doug Labs was able to offer a unique product that empowers clinicians and improves patient care.
Let’s Talk About Building AI That Works in the Real World
No vague promises. No pressure. Just a real conversation with our technical leadership to understand your needs and explore the best next steps.
What Happens Next?
- Your message goes straight to our core team – not a sales department.
- You’ll talk directly with our CEO, CTO, or a domain expert for a consultation-style first call.
- We’ll ask the right technical questions to understand your challenge and scope.
- If we’re a fit, we’ll follow up with a proposed path forward.
More Projects Like This
Pardigm
Computer Vision
Edge AI
Hormone tracking application powered by Computer Vision with 93%+ accuracy in real-life conditions.