Trading intelligence
Natural Language Processing project to parse and analyse human trading information.
02 Services
AI /
ML
dotNear helps customers implement Artificial Intelligence & Machine Learning in their business initiatives. This includes enhancements to clients’ proprietary AI/ML products and developing and training new complex machine learning models.
Excellence center
From data
to deployed models.
- Scope
- Data prep → deep learning
- Platforms
- AWS · Azure · Google
- Projects
- NLP, CV, recommenders
Most commercially successful software applications are beginning to include artificial intelligence — from bots and online sales to algo-traders and other services, where past and general behaviour can provide a better experience for each user. Much of the Artificial Intelligence used on the market right now was developed with the help of nearshore AI/ML teams — a natural result of the long-cultivated mathematical skills found in Central and Eastern Europe in general, and very much so in Romania.
dotNear can provide nearshore AI/ML engineers with expertise ranging from data preparation (data scraping, data wrangling, cleansing and pre-processing) to deep learning architectures (deep neural networks, deep belief networks, recurrent neural networks). We can help build, train, deploy and operate machine learning models fast.
Languages, libraries and platforms
Languages
- Python
- C / C++
- R
- Java
- C#
Libraries & engines
- TensorFlow
- Scikit Learn
- PyTorch
- Keras
- KNIME
Platforms
- Google AI Platform (Unified)
- AWS Machine Learning / SageMaker
- Microsoft Azure Machine Learning
Scope
- Model selection
- Text mining & image mining
- Association rules mining
- Data preparation
- Data preprocessing & visualization
- Data loading & transformation
- Classification
- Clustering
- Dataset enhancement
- Dimensionality reduction
- Regression
- Recommenders
- Distributed linear algebra
- Distribution
Previous machine learning projects
A selection of the work our AI/ML engineers have delivered inside client teams.
Sales prediction
Prediction algorithms based on historical data, deployed into the client’s planning process.
Predictive maintenance
Failure prediction for industrial equipment using sensor data and time-series models.
Image recommendation
Deep learning and computer vision over large sets of images, assisted and unassisted.
Social recommendation
Filtering and recommendation for a social network with a large, active user base.
Driverless systems
Computer vision components for driverless car systems.
Virtual assistants
Multi-intent language interpretation for various virtual assistant products.
Next step
Send us a project brief. Or just say hi.
A short note is enough. We reply within one business day, and a qualified shortlist can follow in 48 hours.