ATLANTIS

1.- Project context. The objective of ATLANTIS is to generate software code for RPA robots in an automated way, based on a (or several) video recording of the sequence of the steps that you want to automate with that RPA. To do this, we will integrate various Generative AI model-agents into the processing flow or pipeline, to interpret the video, decompose it into events, generate the software code for each event, and evaluate said code. We will also improve this processing flow iteratively using adversarial agent techniques. 2.- Topics. #adversarial agents · #generative ai · #llm large languaje model · #lm languaje models · #lmm large multimodal model · #rpa robotic process automation · #intelligent automation · #video interpretation. 3.- Main goals. Improve the production chain and start-up life cycle of intelligent business process automation solutions: Optimizing costs building and deploying intelligent automation proceses / programs; Enhance developer productivity; Reduce time-to-market building and deploying RPA proceses / programs; Reduce the need of involving business key users when taking specifications for building and deploying RPA proceses; Ensure coverage and traceability of requirements implementing and deploying RPA proceses; Optimize energy efficiency in software production and in the deployment and maintenance in execution of said software, etc. And with this, enhance the value provided in broader scope services such as BPO, industrial automation, real-time process monitoring, etc.