Leave a Review
0 Companies  in Shortlist
0 New messages

Change History Through Innovation.

5.0 24  reviews
Silver VERIFIED

CHI Software is a software development service provider operating on the global IT market for 16 years. Our company has representatives and development centers in different parts of the world: the US, the Republic of Cyprus, Ukraine, and Japan.   

$50,000+
$50 - $99 / hr
250 - 999
Founded 2006
3 Languages

3 Languages

  • English
  • Japanese
  • Ukrainian
L'viv, Ukraine
Map location
headquarters
other locations
  • 8270 Woodland Center Blvd
    Tampa, FL 33614
    United States
  • Isogamidoori , 4-1-14
    Kobe 651-0086
    Japan

Focus

Service lines
30%
20%
15%
15%
10%
10%
Custom Software Development
Artificial Intelligence
Mobile App Development
show all +
Client focus
65%
20%
15%
Midmarket ($10M - $1B)
Enterprise (>$1B)
Small Business (<$10M)
Industry focus
20%
12%
12%
12%
11%
11%
11%
11%
Financial services
Business services
Information technology
AI Focus
34%
33%
33%
Application Platforms
30%
30%
20%
15%
5%
Frameworks and CMS
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
Programming & Scripting
10%
10%
10%
10%
10%
10%
10%
10%
5%
5%
5%
5%

Portfolio

Key clients: 

Minespider, MediaMarkt, Sabre, Vodafone, PrismTech, Dressler Consulting, Cyren, BTO, ClimaCell

Parcel tracking solution

Parcel tracking solution

Business needs  

The main idea of the project was digitalization and automation of manual labor, in order to optimize resources, as well as the speed of processing postal information. The initial solution involved much manual human work, which cost too much and was time consuming. 

Bar codes on parcels often get mechanical scratches or other damages before reaching the sorting point. Such defects obstruct the correct identification of a parcel in the system, causing delivery mistakes.  
The client needed to prevent possible mistakes caused by damaged bar codes and also reduce human-made mistakes. 

Provided solution   

CHI Software team started work on this project validating the client’s idea and proving the concept. The main tasks were detection and recognition of objects and image preprocessing. We also selected and trained neural network models to ensure that parcels, letters, and other items of interest are identified automatically and sorted correctly.  
We were responsible for the backend part of the project. Our design team was only involved in minor tasks connected with the UI for parcel station operators.  

Product features 

  • The automated parcel delivery processing system allows: 
  • Correct detection of the postal label among many others; 
  • Extraction of country and zip codes and text if the bar code is not readable; 
  • Compositing the delivery information; 
  • Checking if the postage information is immediately transferred into the database; 
  • Creating instructions for automatic sending of the parcel to the proper vehicle for further delivery.

Technologies:  

Tech stack: Python, Optical Character Recognition (OCR) with Tesseract, С++, NumPy, Pillow, OpenCV, scikit-image, TensorFlow, Keras, CNN, RNN, Darknet, Yolo, CNN-RNN-CTC

RPA robotic process automation for logistics

RPA robotic process automation for logistics

Business needs  
  
The client’s company is spreading its distributed network of contractors to the USA, Mexico, and European countries and needs a customized RPA solution to process a lot of logistics documentation.    
  
Every set of docs usually has from 10 up to 25 documents in different formats and languages. The customer company had a huge human staff to process these documents manually. In 2020, when the pandemic started, it became a necessity to reduce human staff to reasonable numbers and cut business costs with the help of RPA process automation. 
   
The customer was interested in custom RPA solutions to handle high-volume, repetitive and time-consuming tasks with documents and emails that required a large human staff.   

Provided solution  
  
CHI Software delivered an RPA solution based on Machine Learning, Natural language processing, Computer Vision, and Optical Character Recognition. This solution replaces manual human labor in the general workflow.   
  
The RPA process automation system allows to receive separate incoming logistics documents, upload one-page documents as scanned images, determine their type by checking the established set, extract text from predefined page fields, detect the signatures (if they are provided), recognize the QR codes, and compare data entered earlier in their system. In this case the system returns them with request/response, through the API.  
  
This RPA robotic process automation solution also allows you to expand existing data (add new documents and recognizable fields to them, as well as edit existing documents).  
  
We are also planning to make an application for sorting images of documents and extracting information from them. We will also create a service, hosted in Azure, for automatic reading and processing of email messages and a web application for sorting the results of processing.   

Technologies:

Python, Angular, OpenCV, Tensorflow, PyTorch, NumPy, Neural Networks  

Digital payment gateway

Digital payment gateway

Business needs  
  
Development of a digital payment gateway that interacts with existing and future devices for registering and making payments, allowing the end customer to provide conditions for choosing and paying for various services.  
  
Business/tech challenges

  • Innovative online payment gateway solution from scratch that connects the customer’s mobile app, mobile cashier app, POS, and payment service(s);   
  • Compliant with the international security standard rules, PCI DSS;   
  • Changing requirements and features;   
  •  Constant need for new development and support resources;   
  • Integration with numerous payment systems, including MasterPass, ApplePay, and GooglePay.   
  • Change requests realization. 

Provided solution  
  
The implementation assumes the use of the Azure microservice cloud architecture.This project is considered as an element of the creation of the Customer's digital ecosystem.  
Our team was given the task of creating a Virtual Point of Sale, which represents a physical Object virtually and maintains the states of this Object. As a part of the solution, we needed to develop a payment service as a separate module, compliant with PCI DSS.  
  
The external payment service system should be able to provide online payments service for users. This project is in the process now.  

Product features

  • Tokenized, encrypted payments support for enhanced security;   
  • PCI DSS compliant;  
  • Easy integration with the most demanded payment systems including MasterPass, ApplePay and GooglePay, and NFT (in the upcoming product version);  
  • Fast processing speed;   
  • Multiple payment options;   
  • Processing swiped or dipped transactions at a point-of-sale system (POS) or on a mobile device.   

Technologies:
   
Node.js, React.js, C#

Azure Microservices: Key Vault, Service Bus, Cosmos DB, Data Lake.

AI-based vehicle health management system

AI-based vehicle health management system

Business needs

Modern vehicles are complex machines that are costly to fix and maintain. For instance, warranty claims, recalls, maintenance, and vehicle downtime costs reduce commercial fleet and customer satisfaction.

The main business needs were:

  • to allow users to compare and evaluate different anomaly detection models on unsupervised data;
  • to enable faster selection of anomaly detection model for every new client dataset with a few labeled anomalies;
  • to develop a robust data augmentation framework to help recognize unknown types of anomalies;
  • to find a development team with good communication skills ready to start fast;
  • to strengthen the in-house development team with skilled ML developers and Data Scientists.

Provided solution

  • Using Deep Learning technologies, the client’s solution can determine the vehicle's health conditions and then flag any deviations from the expected behavior. This VHM solution can also predict when a vehicle's health condition could lead to failure and help identify the cause of the problem.
  • Unlike other similar VHM solutions that can only operate in the cloud, this one can operate onboard vehicles in real-time, assisting fleet owners to predict vehicle health issues.
  • CHI Software Data Scientist is currently working on implementing and testing Deep Learning models for unsupervised anomaly detection and auto-ml solutions based on synthetically generated anomalous data along with research on related state-of-the-art models and novel ideas in this field.
  • The tasks of the CHI Software’s Machine Learning Engineer include obtaining models from Data Scientists, running them on Big Data, and parallelizing calculations on the cloud.

Technologies:

Python, Microsoft Azure, Databricks, PySpark, Pytorch, Keras,Horovod, Matplotlib, Pandas, NumPy, Sklearn, Seaborn, Plotly, Scipy

3D Head Reconstruction with AI Approach

3D Head Reconstruction with AI Approach

Business needs

The customer’s agency had 100+ in-house 3D designers to process every image.

In the past, the flow used to involve designers processing images from clients. The pictures could be poor quality, from social networks or family albums.

Then designers took stock prop of heads, hands, torsos, etc., and sculpted a 3D model that suited the sent image as much as possible. The image was overlaid on top of the created model as a texture. Every image took 15 minutes to process.

Provided solution

We did not take data from designers because they considered the visible side of the head only, thus their 3D models were not symmetrical and had a non-constant number of vertices.

We had to use something other than the state-of-the-art solutions like 2D to 3D neural networks because:

  • we were missing a proper dataset;
  • the neural network had a huge size and an enormously long training time;
  • even the state-of-the-art solution had quite low-quality results for the provided examples.

Product features

  • Take any random 2D photo shot as an input;
  • Detect the face with RetinaFace deep face detection library;
  • Estimate the rotation with FSA-Net neural network;
  • Find 2D landmarks with FAN2D network;
  • Estimate the scale;
  • Estimate PCA coefficient for Nvidia ICT-FaceKit head models;
  • Create a full 3D object that looks similar to the original face on the photo and is properly scaled and positioned.

Technologies:

TensorFlow, PyTorch, Nvidia, MXNet, OpenCV, Python, Numpy

Health&Fitness Mobile app

Health&Fitness Mobile app

Business needs

During the pandemic period, there was a huge problem with social interaction, since it was not recommended to gather in groups. The industry of fitness apps needed a new solution to cope with social distancing and support the competitive spirit.

Provided solution 

We came up with a solution to create a so-called Leaderboard by location.

The main idea behind this feature is the following:

  • A user could enter the city/state competition table, as well as in the country-level competition.
  • It was possible to compete in different sports.
  • Each company (employer) had its own Leaderboard, competition, and leaders.
  • People don't just do exercises but also earn points for the overall score to compete and motivate each other.

Product features

  • User personalization. Collect user info such as age, gender, weight, height, and so on for customized nutrition and/or workout plans.
  • Wearable devices integration.This functionality allows synchronization of data across devices.
  • Daily/Weekly/Monthly results. This feature helps evaluate the progress in a given period and set long-term goals accordingly.
  • Setting and tracking goals. Aiming at measurable results, the fitness app’s main option is to set the desired results.
  • Overview of core metrics. Sport is about numbers and results, so everything is countable in the app: sets, calories, hours, kilometers, kilograms, etc.
  • In-app notifications. This feature helps improve the engagement of the app users, and overall time spent in the app.
  • Share results to socials. Integration with social networks is a must-have feature for fitness apps.
  • In-app community. To add competition to the fitness routine we added a feature to build a community of like-minded people within the app.

Technologies:

Language: MVVM + feature module architecture Java, Kotlin DI: Dagger2, Coin Multithreading: RxJava2, Coroutines DB: Room, Sqlite, FB DB, FB analytic, FCM Jackson, Moschi, Zxing, Google JetPack, Retrofit, WorkManager, CustomViews

Hail Scanning System

Hail Scanning System

Business needs

Hail is one of the most unpleasant weather conditions, which makes much damage to cars. Detecting it takes time. Today, drivers report hail dents manually, leaving a large margin for human error. Since checking the reported dents manually is time-consuming, recovery price and time are not clear at the moment when the incident is reported. Our client was looking for a software solution that eliminates the need for manual review, delivering the exact count of hail dents with an accuracy rate of 80 percent or higher, as well as decreasing processing time.

  • Our client needed software to inspect the hail damage, identify all types of vehicle damages, and also decrease processing time since checking dents manually is time-consuming.
  • Our client wanted to improve the vehicle scanning system that the industry uses for handling hail damage and make it faster and more accurate.
  • Our client was looking for an effective solution within 0.5 months.

Provided solution

The solution provided by the CHI Software development team detects all hail damages on vehicles, including minor scratches. The detecting process takes approximately 1 minute in total.

Thanks to commercially reasonable resources used by the CHI team, the solution was developed on the shortest possible timeline, with the necessary quality, safety, and business standards integrated.

Here are the details of what the process of detecting damages looks like:

  • A vehicle passes through a light tunnel equipped with optical cameras;
  • High-resolution cameras and the software identify every dent and generate a report based on the damage level of the vehicle;
  • The scanned data about the condition of the vehicle is compressed and accumulated in safe cloud storage.

Technologies:

Python

Artificial neural networks: YOLOv3, MobileNet

Computer Vision: OpenCV, Keras, TensorFlow in Python to build CNN - RNN networks, HOG, Gabor filters 

Image Recognition Service

Image Recognition Service

Business needs

In Japan, many people send their children between the ages of 0 and 5 to preschool, where kids stay on weekdays from morning until 3-6 pm. The conditions for children in preschool institutions are monitored through regular reporting established by the institutions.

The goal of the project was to help educators to use face recognition technologies in the following:

  • Simplify the interaction of educators with the recipients (parents) of the child depicted in the photo.
  • Display the previous number of photos and the photos of each child sent previously.
  • During the shooting, mark the children whose photo it is forbidden to take.

Provided solution

CHI Software development team was involved from the very beginning. We conducted research, tried, and tested more than 10 neural networks for obtaining landmarks and face embeddings. Most of them had limitations that didn't allow us to run them locally on the device. But at the end of the day, we succeeded.

We solved the following issues:

  • Registration of children (the ability to take a photo of one child and save it to the database on the device)
  • Processing of the faces in the photo in real-time
  • Comparing the detected face images with the face images in the database and labeling the found faces
  • We had limitations on the accuracy of the F1 measure not lower than 85% and the speed of the system not exceeding 3/4 of a second

Client values

  • In the POC phase, we developed functionality for registering children and searching for faces in real-time.
  • We managed to achieve the processing and recognition of 5 faces at the same time and the speed of 3/4 second in real-time.

Technologies:

Libraries: Pytorch_android, Pytorch_android_vision, Tensorflow-lite-vision, Tensorflow-lite-meta, kotlin-stdlib

Models: yolov5n — face detector, mobileNetV3_face_lands— face landmarks, moibleFaceNet_m1sm — face recognition.

Android: ML - PyTorch mobile, TensorFlow Lite, DB – Room, Multithreading - Coroutines

Recommender System for the Social Network

Recommender System for the Social Network

Business needs

Geosocial networking apps allow users to share their locations as well as find recommendations for locations or 'venues'. Sharing location has great advantages: it can help us find our way, pick restaurants and shops, and even locate nearby friends and other people within the selected radius. Our client wanted to create a unique geosocial networking solution powered by AI.

The client’s main idea was to change a common approach to social networking apps, add voice and language recognition and create a unique recommendation system

Provided solution

Our team was in charge of the project, starting with the discovery phase and POC, and finishing with the development of the app prototype.

  • This social networking app is organized according to the principle of calendar flow. It is formed depending on the time and location of the app user. The content is broadcast from the place where the user is located and connected to the Internet.
  • The app automatically recognizes the event, date, time, and location of the user thanks to Named Entity Recognition ("What", "Where", "When") or voice assistant. Based on this information, the user receives a list of recommendations.
  • Social communication is possible through the internal chat platform with the functionality to post photos and video files, and comment posts of other users.
  • To cope with a large number of user data, our ML experts have built a knowledge graph of the social network. This custom database helps save and structure different facts about clients for further analysis.

Technologies:

ReactJS, JS, Adobe tools, Amazon AWS, CSS, HTML, REST API, Google Maps, AWS, Postman, Yarn, Adobe Illustrator, HTTPS, Babel, ES5, Redux-Saga, Jenkins, ES6+, Bash, BEM, Axios, Webpack, SASS/SCSS, React markup, Redux, AJAX, Browser Developer Tools, CI/CD, ES6, Google API, Canvas, ES8, JSON, Zeplin.

Python, PyTorch, Pandas, sklearn, S3, EC2, SQS, ML, OpenNLP, Elasticsearch, Flask, Spacy, NLTK gensim

Auto insurance solution

Auto insurance solution

Business needs

It is widely accepted that cars begin to experience a few issues as they get older. Regular replacement of parts, oil, and other components keeps the car mechanism working at a good level. If fuel consumption increases, it's an indicator of a serious problem with the vehicle.

Our client, a taxi fleet owner, noticed a relatively fast deterioration of car parts in one insured fleet. Damages led to extra expenses on insurance, maintenance, and repair services.

  • Сollect real-time data to analyze the causes of rapid deterioration, road conditions, and fuel consumption;
  • Set the smart coverage for aggressive drivers and help regular drivers save on auto insurance;
  • Create a solution that reduces the likelihood of a road accident.

The solution had to use Secure Telematics Hardware with an OBD II interface and be fast and cost-effective .

Provided solution

CHI Software developed the Connected Cars X — a software solution that collects required car parameters with the help of IoT devices. The app uses AI for processing real-time and aggregated data to get behavior analytics.

We built our software solution on cloud computing and Machine Learning technologies. We used an OBD II connector, available in all modern cars, to collect data.

While driving, the RSA encrypted data is sent to the cloud every second via MQTT secure protocol.

Product features

  • Connect OBD II connector. Receive, save, and transmit data to the cloud via the MQTT +SSL protocol;
  • Calculate the car’s acceleration, harsh braking, average speed in general and while entering a turn;
  • Define the average driving performance by hours, days, months, with the ability to view details;
  • Identify segments of the way where driving doesn’t meet the specified safety criteria;
  • Analyze the driving performance.

Technologies:

Scikit-learn, SciPy, FastDTW, FilterPy, Python, C++, JavaScript, AWS IoT Greengrass, AWS Kinesis, AWS Glue, AWS Athena, MQTT +SSL, OBD II

Customizable Travel Insurance App

Customizable Travel Insurance App

Business needs

Our client, one of the UK’s well-known UK tour providers, had a high-load travel web platform with numerous end-users. The client wanted to create a travel insurance app to cover all platforms and especially mobile devices. The main idea was to provide the web platform users with a new mobile experience and allow them to issue an insurance policy, claim it or cancel travelers’ insurance on the go.

  • This client needed a real-time insurance app from scratch, starting with a business process of obtaining insurance for a tourist and scripting many possible scenarios.
  • The client was looking for a professional dedicated development team to offer the best travel insurance solution.
  • Our team suggested cross-platform app development on Flutter to cover all platforms and save clients time and money on native apps development.
  • The client also needed an experienced Product Owner to manage the remote teams from Great Britain, Belgium, Argentina, and Ukraine in close cooperation with the business owners.
  • Finally, the client was looking for effective integration of his travel insurance solution with popular payment systems.

Provided solution

The client’s team had many ideas for travel app features, so the most effective way was to organize our work according to Agile methodology with 2-week sprints.

  • Our BA and QA team gathered, analyzed, and processed many user stories from a UX perspective and insurance policy requirements.
  • CHI Software development team also had an in-house Project Manager. He was in charge of all distributed teams and took an active part in product development.
  • Our development team successfully completed the minimal viable product (MVP) design and development and presented a working prototype to our client.

Technologies:

Node.js, Dart/Flutter, Ingenico payment gateway integration, AWS S3, AWS EC2

Web Portal for an Insurance Company

Web Portal for an Insurance Company

Business needs

The client’s main idea was to provide a unique SaaS platform for insurance companies and their partners. This insurance web portal should cover the full insurance process – from creating insurance contracts to handling insurance events and managing accommodation.

Since the client wanted to enter the market as soon as possible, the main requirement was to create the first stable MVP version in 2 months.

  • Absence of an appropriate web platform to communicate with a nationwide telco partners and retailers;
  • The client needed an insurance app that covers full nationwide document management. All communications among POS, agents, clients, and the company’s employees were separate and segmental;
  • Providing a simple sales process by automating the calculation of value and registration of an insurance contract;
  • Unique MVP design and development of the stable product version within a short period of time;
  • Attracting new customers and offering high-quality services to existing ones through the usage of a single customer base and CRM software solutions.

Provided solution

The CHI Software team has developed an insurance web portal focused on managing both B2B and B2C processes within the insurance company. The solution provides a platform to organize and track interactions with clients and partners, both current and potential.

The web platform includes a list of necessary functions and paperless workflow instruments. The tool for signing all documents by an electronic key was also added.

Since the customer was looking for a unique system design, we could not use out-of-the-box libraries for solution interfaces. That’s why many of the insurance app components were either modified or made from scratch.

Technologies:

Java 11, Spring, Hibernate, Liquibase, PostgreSQL, JODConverter, RabbitMQ, React JS, TypeScript, Redux, Redux-Thunk, Ant Design, Less, AWS (ECS, RDS).

Reviews

5.0 24 reviews
Submit Review

Verification

Clutch verification provides an additional layer of data to help you make the right purchasing decision of business services. Learn more
Silver VERIFIED
Business Entity
Business Entity Name
CHISW DEVELOPMENT LTD
Status
Active
Jurisdiction of Formation
Cyprus
ID
HE372271
Date of Formation
July 31, 2017
Source
Republic of Cyprus, Ministry of Commerce, Industry and Tourism
Last Updated
July 21, 2022
Payment & Legal Filings
International Credit Risk Assessment Per our methodology, the International Credit Risk Assessment rating is determined by our analysis of a credit report and third-party data.
Moderate Risk
Source
Creditsafe
Last Updated
July 21, 2022
Client Reviews
VERIFIED CLIENT REVIEWS
23
OVERALL REVIEW RATING
5.0
Source
Clutch
LAST UPDATED
November 14, 2022