System Development with artificial intelligence (AI)
Efficient Problem Solving with Machine Learning (ML) Models
Preparing for AI
3 successive stages of work
- Description of the idea in those document
- Machine learning problem setting
- Documenting the expected result
The TOR describes: what needs to be achieved with the AI and additional components of the system, and where the input data will be taken from, where it will go next (intermediate processing) and in what form and the final result of the system will have to be displayed (where to integrate later) raquo;.
from $ 45 000 /
from 10 days
Table of contents screen from TK
Find by Face project
Screen from analysis
Find by Face project
Screenshot of a mind map from the
architecture of the Find by Face project
Create AI
Using ready-made models
- Model setup
- Data preparation
- Model training
- Evaluation of the result
- Optimization
The creation of artificial intelligence takes place in several repetitive stages. Adjusting the finished model to the requirements from TK. Data collection (existing data is structured, search is carried out among open sources or created from from scratch). Model training (i.e. training to get the desired result from it). Evaluation of the result (testing on data that the model has not previously seen). Optimization (new data, training, evaluation).
from $ 675 000 /
from 45 days
AI system
Implementation of the entire system with AI on a turnkey basis
- Preparation (TOR, Analysis, Architecture)
- Creating AI (open source or API model)
- Implementation of related components (Back-end, DBMS, admin panel, API)
- Integration into business (existing IT system)
An effective solution to the problem consisting of implementation of a set of stages aimed at achieving the desired result. The preparatory stages are carried out taking into account the existing IT system. The creation of AI is carried out in parallel with the development of related components. The system is created functionally scalable and easily supported.
from $ 1 575 000 /
from 120 days
- AI development
- Ecosystem for AI
- AI functionality
AI development
Platform selection, model creation and training
PyTorch, PyTorch Lightning, Fast.ai, TensorFlow, OpenCV
Machine learning from Apple / Google, MTCNN (facenet-pytorch), InceptionResnetV1 (VGGFace2), BERT DeepPavlov, ruGPT-3
Plate Recognizer, Amazon SageMaker, Amazon Vision, Yandex DataSphere, Yandex Vision, Azure Computer Vision, Azure Decision, BERT from Google
Ecosystem for AI
Bundle of back-end and mobile application
Figma, After Effects
MVVM for iOS with Combine and SwiftUI
MVVM for Android with Kotlin and Android Studio
Python, Django, Docker, Swagger
PostgreSQL, MongoDB
Chat (tete-a-tete, group), Invites, Subscribers, Subscriptions, Rating, Reviews, Complain
Trello, GitLab, Telegram, Google Workspace
AI functionality
Artificial intelligence can do a lot
Object Detection, Defect Detection, Object Counting, Anomaly Detection, Anomaly Detection, Pattern Detection
Predicting an event / action / disease, Forming recommendations, Revealing patterns, Searching for anomalies
Recognizing words from «Audio/Video/File/Speech» and then converting them to text
Determining the sound source, its characteristics or composition
Implemented projects
Steps implementations AI
- Functional requirements
- Terms of Reference
- Analysis
- architecture design
- Create AI
- Packaging the AI into a module
- Integration of the AI module into business
- Quality Achievement Control
Contact Service
by Develop Marketplace
Form Submit
Submit your application
