Custom and Affordable AI Solutions for Business

Machine Learning | AI | AR | Deep Learning | GANs | Computer Vision| NLP

 

Who are we?

A team of full-time top-level Data Scientists with Ph.D.-level expertise in machine learning and computer vision. 

Whether you are looking for process automation, predictive analytics, big data analytics, deep learning, or AI software development - we offer custom turnkey solutions that guarantee you peace of mind.

 

Why are we better than alternatives?

We offer upscale server infrastructure, affordable pricing, and a stable team of experienced Data Scientists that work together on projects. 

We have a 100% success rate because our world-class Data engineers and infrastructure architects handled the most complex projects with flawless delivery.

Our Expertise: 

  • Computer Vision: models for image classification, object detection and tracking, OCR

  • Natural Language Processing: development of ChatBots, Text Classification, Semantic search and analysis, sentiment analysis, etc.

  • Time Series Forecasting Models: time series prediction models for prediction of energy consumption and automated trading platforms.

  • Generation of synthetic data sets.

Our Projects

We specialize in training models for image classification, object detection and tracking, facial detection and recognition, license plate recognition, detection of fakes, and OCR. Our successfully completed projects include:

  • Facial Recognition, fake face detection

  • License Plate and Vehicle Recognition
  • Automated Monitoring of Solar Panels for preventive maintenance
  • AI-Powered Automated Trading Software
  • Deep Cheque Scanner
  • Data Extraction and Document Parsing Software

 

 
$5,000+
 
$50 - $99 / hr
 
10 - 49
 Founded
2017
Show all +
Rostov on Don, Russia
headquarters
  • Rostov on Don
    Rostov on Don, ROS 344011
    Russia

Portfolio

Automated License Plate and Vehicle Recognition (ALPR)

License Plate and Vehicle Recognition in the Most Challenging Conditions

  • Real-time traffic analysis.

  • Detection and classification of vehicles by type.

  • Accurate reading of license plate numbers.

  • Support of various regional standards of license plates.

  • Capability to aggregate, store, and monitor data from 1000+ cameras.

  • The neural network, trained in Tensorflow, detects vehicles, plate numbers, and letters with digits.

  • Our models are integrated into an optimized, multithreaded C++ agent. It communicates with the server via GRPC protocol. The cross-platform works on Windows, Mac, and Linux. It collects, stores, and monitors data from various camera types.

  • Facial Recognition Software

    Our Data Scientists use advanced techniques of deep learning and computer vision. We achieve high accuracy of face recognition that tackles a variety of business problems.

    We trained a deep siamese neural network to achieve lightning-quick identification of a person. We designed an SDK to deliver maximum performance regardless of input, with an error rate close to 0.

    Heading: Key Features of Our Face Recognition

    Technology:
    • Fast one-shot facial recognition

    • Fake face detection (works for mobile screens and printed photos)

    • High accuracy and security (no false positives)

    • Optimized models via SDK for C++, C#, and Java

    • Solution that works independently with no internet access.

    Image Segmentation using Deep Learning and Computer Vision Image

    Image Segmentation using Deep Learning and Computer Vision

    Example of Image Segmentation Software

     

    Data Scientists at BroutonLab created a Sky segmentation algorithm based on the application of deep learning segmentation methods. The algorithm successfully processes images with trees, buildings, and small objects. The accuracy of the model remains consistent in various lighting conditions

    and can run on mobile devices with low-resolution cameras.

     

    Model

     

    Our model is based on U-Net with ResNet Encoder. We applied a post-processing algorithm to increase the quality and accuracy of the results.

    To train the neural networks, we manually labeled 5000 photos from iPhone and Android devices.

    Basketball game analysis

    Basketball player tracking engine, which helps individual players to improve their gaming skills via automatic statistical analysis of basketball matches.

    Artificial Intelligence Automated Trading Software Image

    Artificial Intelligence Automated Trading Software

    Artificial Intelligence Automated Trading Software

     

    Our data scientists developed software for real-time market analysis. We used deep learning, machine learning techniques, and performance computing to create AI-powered Automated Trading Software. 

    Deep learning can analyze a full trading order log. We developed our

    own data storage and processing algorithms to store the log data, representing over 200 GB of trading information. 

    The trader can adjust and automatically optimize the AI-powered indicator.

     

    Key Features: 

    • Smart Trading signals. A deep neural network analyzes the complete market information and alerts the trader. 

    • It offers intelligent Trading strategies, optimized for such metrics as equity, max drop-down, Sharpe Ratio. 

    • Its visual interface is easy to use to build custom indicators and trading strategies. 

    • OHLCV+ extended trading statistics. Including Order Book, Order Count Book, Open Interest, Executed Volume, Order Count. Statistics are aggregated by levels.

    • The Automated Trading Software is implemented and optimized in CC+ and has its own data storage format. It does not need deserialization of bars and dynamic structures, such as the Orderbook. 

     

    Other Features:

    • Our Automated Trading Software has a scalable architecture. It’s easy to add new connectors for other exchanges

    • User-friendly graphical terminal for manual trading

    • Advanced strategy optimization software

    • API for software developers (Python/C++)

    Intelligent Software for Automated Monitoring of Solar Panels Image

    Intelligent Software for Automated Monitoring of Solar Panels

    Intelligent Software for Automated Monitoring of Solar Panels 

     

    We developed an image processing engine to monitor failures in solar panels.

     

    We used OpenCV and AI-powered object detection techniques to achieve the maximum accuracy of the software. 

     

    Our software can detect anomalies

    and faults by analyzing RGB and infrared images. The images are taken by drones and other unmanned flying vehicles to check the panels in real-time. 

     

    The AI-powered engine timely detects anomalies in solar panels and requires minimum manpower to monitor them in real-time.

    Reviews

    Sort by

    Development Support for Videoboard Film Firm

    "Their analysis work is strong, along with their data science capabilities."

    Quality: 
    5.0
    Schedule: 
    5.0
    Cost: 
    5.0
    Willing to refer: 
    5.0
    The Project
     
    $200,000 to $999,999
     
    May 2018 - Ongoing
    Project summary: 

    BroutonLab developed a facial recognition feature for their app. Their team works an AI model and performs research, too. 

    The Reviewer
     
    11 - 50 Employees
     
    Middlesex, New Jersey
    CTO, Videoboard Film Firm
     
    Verified
    The Review
    Feedback summary: 

    The deliverables from BroutonLab have saved time and people power for the client. They work independently and manage the workload well. They demonstrate deep expertise and are skilled within their space. 

    A Clutch analyst personally interviewed this client over the phone. Below is an edited transcript.

    BACKGROUND

    Introduce your business and what you do there.

    I’m the CTO of a videoboard film firm.

    OPPORTUNITY / CHALLENGE

    What challenge were you trying to address with BroutonLab?

    We needed help with development.

    SOLUTION

    What was the scope of their involvement?

    They implemented an improved method for facial verification in our app. They build the AI models, too. They do good research on which model and how it should be implemented.

    What is the team composition?

    I work with two data scientists and their manager.

    How did you come to work with BroutonLab?

    I found them on Upwork. I interviewed them severely and they survived the test.

    How much have you invested with them?

    We’ve spent $300,000.

    What is the status of this engagement?

    We started working together in May 2018 and the relationship is ongoing.

    RESULTS & FEEDBACK

    What evidence can you share that demonstrates the impact of the engagement?

    They moved the content that we don’t need to process. Originally, we had to process three hours of footage and now that the detection feature is accurate, it’s closer to 30 minutes. They minimized the demands of labor and hours that were needed to complete the task.

    How did BroutonLab perform from a project management standpoint?

    They’ve managed the projects almost by themselves. I typically give them assignments and they perform it. They coordinate the people who work on different tasks. 

    What did you find most impressive about them?

    Their analysis work is strong, along with their data science capabilities. 

    Are there any areas they could improve?

    No, everything is good.

    Do you have any advice for potential customers?

    They research all opportunities because they’re extremely intelligent people. Give them the problem and they will solve it. It doesn’t even have to be extremely specific.

    5.0
    Overall Score I enjoy working with them
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 5.0 Cost
      Value / within estimates
    • 5.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer

    Program Translation for Freelance Ecosystem

    "They're clear and effective communicators."

    Quality: 
    5.0
    Schedule: 
    5.0
    Cost: 
    5.0
    Willing to refer: 
    5.0
    The Project
     
    $50,000 to $199,999
     
    Jan. 2018 - Ongoing
    Project summary: 

    BroutonLab was hired to help a freelance ecosystem translate code for a hierarchical Bayesian model MCMC estimation routine from R/C++ to CUDA.

    The Reviewer
     
    1-10 Employees
     
    San Francisco, California
    Vlad Lokshin
    Co-Founder & CEO, Turtle.Dev
     
    Verified
    The Review
    Feedback summary: 

    Once completed, the project significantly increased the client's computing speed. The company was impressed with the amount of talent the team had at its disposal.

    The client submitted this review online.

    BACKGROUND

    Please describe your company and your position there.

    I'm the CEO of Turtle.dev.

    OPPORTUNITY / CHALLENGE

    For what projects/services did your company hire BroutonLab, and what were your goals?

    The team was hired to help our client translate code for a hierarchical Bayesian model MCMC estimation routine from R/C++ to CUDA.

    SOLUTION

    How did you select this vendor and what were the deciding factors?

    We were referred to Michael by another developer working with us.

    Describe the scope of work in detail, including the project steps, key deliverables, and technologies used.

    The team helped our client translate code for a hierarchical Bayesian model MCMC estimation routine from R/C++ to CUDA.

    How many people from the vendor's team worked with you, and what were their positions?

    3 developers covering Data Science, ML, and backend development roles. There has also been minor frontend work for showing data processing results.

    RESULTS & FEEDBACK

    Can you share any measurable outcomes of the project or general feedback about the deliverables?

    This translation allowed speeding up computation time by two orders of magnitude! This also allowed paralyzing the code onto a high-power NVideo GPUs.

    Describe their project management style, including communication tools and timeliness.

    Michael's team work with our customer through through Turtle's app. They're clear and effective communicators.

    What did you find most impressive or unique about this company?

    Super capable and prices Data Science and ML experts. Just a very talented team overall!

    Are there any areas for improvement or something they could have done differently?

    na

    5.0
    Overall Score
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 5.0 Cost
      Value / within estimates
    • 5.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer