AI-Assisted Code Reviewer:
This AI solution is the first of its kind to automatically identify parts of code that need human review. It incorporates NEC’s leading-edge AI, software engineering and system integration to tackle current opportunities for growth in the software development process. More efficient than current labor-intensive methods, and highly accurate, it achieves 10x higher productivity over manual review. It suggests the most relevant reviewer based on review history and code similarity, and can automatically generate code review recommendations. By maintaining human involvement, this first-of-its-kind code review solution uses continued learning of domain experts’ knowledge for the AI agent.
Multimodal Analytics for Time Series and Text Comments:
Powered by DL, this technology offers data interpretation for automated analytics, particularly time series data. It provides human-readable explanations from time series analytics, helps to search for time series using a text query or a mixed-modality query, and enables deeper understanding of time series data using associated text data. The DL-powered time series analytics are coupled with natural language processing (NLP) techniques to extract domain-specific insights with minimal human supervision. This technology can be used for diverse industries, from an operator report from industrial sensors, to the trading pattern description from a financial chart. It creates novel values for customers who are looking to expand into the uncharted waters of new markets.
Fingerprint of Things:
This project uses image matching technology to capture unique patterns from product surfaces. With it, you can identify and authenticate each individual product from your production line using your smartphone.
This project modernizes traditional causal inference by automating the heavy workload for analysts. It provides models from the data, visualizing DAGs, key factors, root cause, and provides predictions and simulations.