How we work
Our approach is Human centric + Data driven with constant focus on research and development
Methodology
Our methodology is based on a consultancy approach capable of working alongside our clients at every stage to identify and implement paths to make business processes more efficient, controllable and automatable.
- Identification of area of intervention
- Definition of the team
- Executive sponsorship
Elif Lab + client team
- Feasibility assessment (go/no go) with respect to: available necessary data (internal and any external sources), processing time, functional requirements
- Feasibility assessment report (macro indication of time, work plan and costs)
- Detailed formalization of the project with precise indication of time, deliverables and costs
- Project acceptance and kick-off
- Process design
- Internal and external data recovery
- Analysis on subsets of data using proprietary tools
- Strategic evaluation of evidence emerging from data analysis
- Evaluation of data quality and completeness with integration of feedback and recommendations
- Application development on complete dataset through composition and integration of proprietary algorithms
- Integration with data flows and client systems
- Configuration of the application according to technical specifications
- Intermediate releases of the application with AGILE mode
- System tests
- User testing
- Drawing up technical documentation
- Final release of the application
- Detailed process report
- Dataset analysis report
- Application training
A continuous relationship:
We support our customers in the maintenance of the application.
We carry out on-demand and periodic analyses of the implemented process and its results.
Research and Development
At Elif Lab, the research and development component plays a central role and constantly feeds our cutting-edge technological solutions.
We have gained experience with different techniques and models.
We experiment with ideas, hybridising ‘niche’ paths with more established ones, trying to find a mix that is both effective and lean.
We like to understand ‘how things work’, not just make them work.
We bring seemingly distant worlds together to create innovation.
Some of our research and development
Model of the mind
The creation of algorithms inspired by our attempt to formalise models of psychological and philosophical origin is giving us conceptual advantages beyond our expectations.
From the point of view of effectiveness, it provides us with the possibility of making more conscious use of various techniques whose affinities would otherwise not be evident.
Proprietary implementations
Some techniques we use are only available in scientific articles (no open-source implementation): we therefore work on developing proprietary implementations.
3D Reconstructions
To extract significant information from photos we are actively working on experimenting with reconstruction of 3d models.
We create 3D models from collections of images using advanced photogrammetry techniques, but we also experiment with the most modern context-aware neural networks for single image reconstructions.
Logic algorithms
Embedding deep learning techniques into logic algorithms allows us to recognize causal patterns and extract general categories in photo content analysis.