AntWorks RPA, ANTstein, was a robotic automation platform (RPA) that featured a machine learning platform built on fractal science to deploy AntWorks Bots, including QueenBot, processes which perform the automation. The product has been discontinued and is no longer available for sale.
ANTstein needs no code or minimal programming. This makes the assigning tasks to bots much easier without wasting time. ANTstein in very useful in performing rule-based tasks. For some tasks and special cases, human assistance is still needed to make decisions.
ANTstein processes unstructured data , which constitutes 80% of the data in enterprises in various industries. It addresses the limitations of RPA by processing various types of data including structured, unstructured, image, and inferred data accurately and fast.
AntWorks combines data curation and digital workforce management to enable end-to-end process automation capabilities for higher straight through processing (STP).
AntWorks offers multi-tenancy in which multiple bots and human to work on same machine.
Within just 3 months of training, ANTstein is able to deliver 85%+ data accuracy.
Most of the times when performing a task on a remote machine, we have to deploy a screen recorder so that we can debug the process bu actually visualizing the workflow. There should be a remote monitoring tool on the Queen machine.
When automating Google spreadsheets to plot regression graphs, there are cases when some cell has NaN values. ANTstein always sends an error log in the admin machine. It skips the row and we end up with a smaller amount of data. We have also made a logic for predicting the NaN values by finding the median and fillna pandas-python library. Thus, improvement in flexibility is needed.
One should have to be familiar with the user experience of the technology or get certified to work upon. Because the tool is not so easy to learn at one go.
ANTstein is designed to automate the entire business process rather than merely a single discrete task. Leveraging fractal networks, rather than neural network technology, ANTstein is able to use relatively small data sets to train the engine thus enabling faster training, testing, deployment and execution of business processes.
Mostly we use data scraping tools and then take deep insights from the extracted data. Till now, we were unable to perform automation after scraping the data with the above-mentioned tools. But using ANTworks, we can automate the complete process. Our mainstream work process looks like: Twitter Hashtag extraction > Saving the scraped data to CSV > Importing the CSV in Python > Natural Language Understanding / NLTK in python > Creating pos/neg columns > giving positive and negative insights using analysis.
