Our team has advanced skills ofGen AI & Machine Learning
To generative
Our services that use the generative
Mario Rossi
Not Authenticated
Gianni Bianchi
Authenticated
Eleonora Ferrari
Warning
There are a lot of cool framerworks out there like React, Angular, Vue, Svelte that can make your life ....
Use PHP :-D
Technologies
These are the main technologies we use
Ollama
Weaviate
langchain
Bedrock
Llama Index
CrewAI
Autogen
Qdrant
Haystack
Redis Vector
And many other technologies selected according to the specific needs of the project
Machine Learning
Deep Learning
We design and build neural networks of Deep Learning based on your needs
Ai for recruiting
Our platform that allows you to analyze a large quantity of curricula and improve the KPI of recruiting
Find out here
Use Case
Autonomous submarine robots for Field Offshore supervision
Develop autonomous submarine robots with AI by inspection and monitoring of underwater oil fields, reducing operational risks and maintenance costs.
Advanced computer vision system with 3D CNN for recognition obstacles and object detection in submarine environments. Implemented edge computing with Nvidia Jetson Agx for real-time processing, integrated with autonomous navigation systems and underwater lidar sensors. Deep Learning Models trained on specific datasets for recognition pipeline, valves and offshore structures.
Autonomous robots revolutionized offshore inspection operations, eliminating the need for divers for 90% of routine inspections. The Ai Predicative system identifies potential structural problems before they become critical, drastically improving operational safety and reducing emergency intervention costs.
Technical architecture
- Computer Vision with 3D CNN for Object Detection Sottomarino
- Edge Computing Nvidia Jetson Agx for autonomous processing
- Ai navigation system with Slam adaptive algorithms
- Deep Learning Models for recognition offshore structures
- Satellite communication for real-time telemetry
- Mlops Pipeline for Continuous Learning from missions
Results
- 95% Reduction Requirement Ski interventions
- Early identification 78% structural anomalies
- 48h operational autonomy without human intervention
- ROI of 210% in 24 months for offshore operations
Use Case
Automatic classification system scientific articles chemical
Automating the classification and categorization of scientific publications in the chemical field for international academic magazines and databases.
Implementation of Transformer-Based architecture with Bert Fine-Tung models on specialized chemical corpus. Advanced NLP system for chemical entities extraction, semantic analysis and multi-label classification. Integration with chemical knowledge graphs and molecular database for automatic validation content and correlation suggestions.
The system has completely automated the editorial workflow of 12 international scientific journals, processing over 15,000 articles/year. AI automatically identifies emerging trends, suggests appropriate reviewers and detects potential conflicts of interest, significantly accelerating publication times while maintaining high quality standards.
Technical architecture
- Transformer Models Bert Fine-Tung on Chimico Corpus
- NLP Pipeline for extraction entities and molecular relationships
- Knowledge Graph Integration with Pubchem and Chembl
- Vector Embenddings with QDrant for Similarity Search
- Langchain Orchestraction for complex workflows
- Mlflow for Model Versioning and Performance Tracking
Results
- Accuracy classification 94.2% on 150 categories
- Reduction 80% preliminary review time
- Automatic identification 67% emerging trends
- Processing 15k articles/year without manual intervention
Contact us
A quote, a particular request, or simply a coffee to meet us? Write to us and we will reply to a "nano-second".