Periodic Reporting for period 1 - DC-IR (AI based software platform)
Periodo di rendicontazione: 2019-06-01 al 2019-09-30
This Phase 1 project consists of studying the viability of the project to develop the DeepCode automated, artificial-intelligence-driven code review tool. The objectives of this feasibility study were to finalise the technical requirements and define the development roadmap for DeepCode, to analyse the market for our solution, ensure regulatory compliance, identify risks and elaborate a business plan for the project.
We have analysed the progress of our work on the prototype and beta version of our service. On the basis of this, we have defined the requirements for a minimum viable product for the next iteration of DC-IR. This culminates in the preparation of an outline of the work plan for the next stage of the project.
On the commercial side, we have analysed the market in detail, which provides inputs to our commercialisation strategy for a successful launch of the professional offering of the service. We have also considered the IPR issues and regulatory issues we potentially face in the exploitation of the project, as well as other risk factors. Finally, we have examined the financial feasibility, preparing financial projections for the company and determining the funding requirements to execute the next stage of the project.
On the commercial side, we have analysed the market in detail, which provides inputs to our commercialisation strategy for a successful launch of the professional offering of the service. We have also considered the IPR issues and regulatory issues we potentially face in the exploitation of the project, as well as other risk factors. Finally, we have examined the financial feasibility, preparing financial projections for the company and determining the funding requirements to execute the next stage of the project.
DeepCode is unique in its ability to apply machine learning to improve the quality of computer programming code. It is able to identify and recommend programming best practices that are embedded in millions of lines of code in public and private software repositories, but not necessarily documented or easily identified. DeepCode’s machine learning algorithms extract those learnings automatically, without reliance on human defined rules, and then used to help developers optimise their code.
The automation of this process dramatically reduces the cost of software development by reducing the number of bugs in code, identifying them earlier in the programming process, when they are less expensive to correct. This reduces the burden on testing later in the process. For software users, this results in better, more efficient, and more secure applications.
The automation of this process dramatically reduces the cost of software development by reducing the number of bugs in code, identifying them earlier in the programming process, when they are less expensive to correct. This reduces the burden on testing later in the process. For software users, this results in better, more efficient, and more secure applications.