Gleematic Functions

Gleematic is Cognitive Automation software that helps you simplify your work with a combination of Artificial Intelligence and UI-based / rules-based automation. Gleematic is flexible to be used in various departments of multiple industries, easy to set up and no coding.


FunctionsWhat it isWhy it is importantWhere to use itGleematic A.I. versionGleematic Basic version
Natural Language Processing – Built in
Sentiment AnalysisThis is a type of artificial intelligence (A.I.) to identify and categorize opinions expressed in a piece of text, to determine if the writer’s attitude/ feeling (be it in general or towards a particular topic) is positive, negative, or neutral.To understand the human’s emotion/ opinions/ attitudes through the written text.It can be used to understand opinions of people in various scenarios. Examples:
HR Department: employee satisfaction
– Customer Service Department: customer satisfaction
– Marketing/ Sales: customers’ opinion on product/ service

Text ClassificationThis is a type of technology to classify text into organised groups. For example, cat, dogs, sheep, can be classified under “animals” whereas carrot, cabbage, spinach can be classified under “vegetables”. artificial intelligence (A.I.) can be applied to make text classification more sophisticated to classify bodies of text, such as determining whether an email from customer is a complaint or a compliment.Conditions can be set for making different decisions based on overall categorising of the written text. This is important as it is impossible to set up comprehensive rules and “if-else” conditions merely based on keywords, because human/ natural language is very dynamic.Examples:
– Classify incoming emails into relevant categories, such as “job applications”, “customer feedback”, “product enquiry”, etc.
– Categorise/ link together information/ documents so that humans can access the right information at the right time.- Provide quick, automated first-level response to the customers, based on text classified, before escalating to humans.

Train own text classifierThis is an “engine” to allow users to define their own ways of classifying/ categorising the text.Users might have text that is specific to his/her company or industry that needs to be classified that may not be pre-built in the software.It can be used whenever users need to classify the data into specific categories. For example, to classify types of machine-defects, medicine, etc.

Train own text extractionThis function allows users to define their own ways of doing text extractionUsers might have text that is specific to his/her company or industry that needs to be extracted that may not be pre-built in the software..It can be used whenever users need to extract data from various fields in documents (some of which might be very specific to the company)

Date ExtractionThis function allows users to detect/get and even manipulate the date format.This function will allow users to handle changing of date formats automatically.Users can use this function for any date format in various type of file like PDF, Excel, Word,etc. and transform this into another desired format (e.g. detecting 23-Oct-2018 and transforming into 23/10/2018)

Text Pattern ExtractionThis is a feature to extract the structured  text using Regular ExpressionThis function allows users to easily get the pattern from semi-structured textWe can use this function to process  semi-structured text that might need regular expression.

Built-in OCR (Optical Character Recognition)OCR is the recognition of printed text characters by a computer. This involves photo-scanning of the text character-by-character, analysis of the scanned-in image, and then translation of the character image into character codes, such as ASCII, commonly used in data processingWithout OCR, data in images need to be copied manually. That would be very time-consuming and prone to errors.This function allows users to digitise data that was captured as image (e.g. certificates, Identity Cards, etc), so that it can be processed by computer/ machines.

Chinese NER (Named Entity Extraction)A “named entity” is a real-world object (e.g. persons, locations, organizations, etc.), that can be denoted with a proper name.This is the first step for information extraction to happen. It classifies information into different types of named entities.This will help users process data or documents that are in semi-structured or non-structured formats, in Asian languages

Indonesian NER

Vietnamese NER (beta)

Thai NER (beta)

Japanese NER (beta)

Fuzzy data-cleaningThis feature uses fuzzy-logic to clean up “dirty” data. It tries to auto-correct data based on the closest possible data (from the training data-set). Users have to set up probability of matching to determine how closely the words are to be matched.This is needed for cleaning up data when there could be accidental mis-spellings, short-forms, missing spaces, etc.It could help auto-correct data detected from documents (e.g. order-forms, delivery-notes) to match with something that the machine was trained before. For example, the word “invoic” can be auto-corrected to be read as “invoice”.

Machine learning- Built in
Supervised learningIt’s the type of A.I. Machine Learning in which both input and desired output data are provided. Input and output data are labelled for classification to provide a learning basis for future data processing.This will allow user to “teach” the software-robot do the task as desired, based on the inputs provided.This function is mostly useful in info extraction process. It can also be used to do predictive analysis (e.g. to detect potential fraud)

Outlier DetectionOutliers are extreme values that deviate from other observations on data, and thus may be an “abnormality”.This will allow users to intelligently classify cases that are not in the “normal range”.This will be useful to detect potential abnormalities as malicious activities or fraud or mistakes made by humans.

Choice of machine-learning algorithmsA variety of algorithms are available.This allows advanced users to decide what is more suitable to be used in specific types of data setsBeginners can use the default settings, and advanced users can have greater flexibility.

Computer Information Extraction
Information Extraction from different types of documentsThis is feature to make the computer recognise and understand the text pattern from any unstructured text, and pull out desired data.To extract data that is not clearly labelled (lacks metadata) or  cannot not readily mapped onto standard database fields.This feature can be used to get data from unstructured/ semi-structured text in various types of documents such as invoices, seaway bills, letters of credit, bank statements, etc. For example, you can extract data such as “date”, “payee”, “total”, from an invoice.

Imbalanced data handlingImbalanced data typically refers to a problem with classification where the classes are not represented equally. For example, there might be hundreds of “short sticks” while there are only a few “long sticks”To mitigate errors that might be caused by over-representation of data in certain classes, while under-representation in othersThis will prevent skewed models which would likely always predict “short sticks” (because of higher representation) and erroneously indicate high accuracy.

Computer Vision
Face detectionA feature that allows users to execute the action only when a face detected.Some operations  may require high security, including face detection.This feature can be use in any customer on-boarding processes (e.g. KYC processes).

Face recognitionThis feature can make the computer recognize the face based on the data saved on the server.Some system require a face recognition to do validation.This function can be used to identify the person for authentication or to trigger specific workflows.

Logo recognitionIt similar with “Face recognition” but rather than recognizing the face, this feature will make computer to recognize the logo.Logos can be used to classify documents or to differentiate workflows.This function can be used whenever we need to classify files based on the logos, or to trigger specific workflows.

Train own image recognition + Detecting imagesThis feature allows users to put in their own images for Machine Learning to train the model file.This allows flexibility to users to be able to train for any type of image, even if confidential, without having to bring in external parties.Users can use this function to train on any JPG or PNG file. For example, a retailer can use it to train on different types of watches or bags .

Image/ Icon countingThis function allows the machine to recognize and count the images or objects shown on the screen.This function is important when we need the quantity of the icons as the condition to do specific task.We can use it on websites or desktop application. Every icon shown on the screen can be counted by the software-robot.

UI Automation:
MouseA feature to control the mouse clicks and movements automaticallyThis is one of the main input tools to operate the computerSince this is the main input tool, this function is very useful in every automation process.

KeyboardA feature to control the keyboard strokes automaticallySame as the mouse, keyboard is also another main input tool to operate the computerSince this is the main input tool, this function is very useful in every automation process.

Data entryA function that enables users to type characters or text to be entered into relevant fieldsIt is important because every data transaction will need data entry.It will be useful for many operations such as transferring data through multiple system, info extraction, etc.

Data manipulationA function that allows users to manipulate the data. (e.g. to change the form, to do arithmetic operation, etc)This is important as raw data might need to be transformed before being used.We can used this function to help us in many process (e.g. generating report, loan approval process, etc)

Citrix SupportAn application provides desktop virtualization.It allows users to work with applications such as those requiring Citrix support.It can be used whenever you need to do the operation in virtualization app.

Legacy System supportA feature that make us possible to automate operations through the legacy system (e.g. telnet legacy)Some operations are still relying on the legacy system. We can’t just eliminate that way. We need to create an integration between the existing systems.In every legacy system (e.g. AS 400, DOS, etc)

App integration:
ExcelFeature that allows us to automate the operations in Microsoft Excel. (e.g. copy the list of employee ID from the employee data)Microsoft Excel is the common format that being used to present the data.It can be used every time we need to do the operations with the Microsoft Excel files.

WordFeature that allows us to automate the operations in Microsoft Word.Microsoft Word is one of the common format that being used to present the data.It can be used every time we need to do the operations with the Microsoft Word files.

CSVA function that allows us to automate operations that related to the CSV fileBecause some existing functions are still using the CSV file format.It can be used every time we need to do the operations with the CSV files.

xmlA function that allows us to automate operations that related to XML fileIt’s important to make us easily do the operations in any XML fileThis function will be useful when we have the XML file and we need to get or manipulate data from the file.

HTTP requestsTo communicate with another web based systems.Because many operations in company are done through various system including web based systemThis function can be used whenever we need to call the API

REST API integrationTo communicate using the “REST” protocolThere might be a need to communicate with other appsThis will allow other applications to get connected to Gleematic using the REST protocol

Custom ScriptsUsing programming language to create new functionsThere might be a need to incorporate customised functions.It will allow flexibility to extend the functionalities and customise process as needed.

Invoking PowerShell scriptsPowerShell is an object-oriented automation engine and scripting language with an interactive command-line shell that developed to help IT professionals configure systems and automate administrative tasks.We need a function to invoke the PowerShell script so we can create integrated automations between all the systems.You can use this function whenever you need to execute the PowerShell script.

Error handlingThis is a form of anticipation, detection, and resolution of programming, application, and communications errors.It’s important because we may do mistake while creating the script for the robot.This function can be used in every automation processes.

Accessing important Apps:
Terminal (e.g. AS/400)This function allows the robot to access the terminal (e.g. AS/400).It’s important especially for the company that still using the legacy system.We can rely on this function for existing legacy system.

OutlookThis feature makes the robot possible to do interaction with the email application such as Outlook.Because email is the main component to communicate with other parties that related to the company.We can use this function whenever we need to manage the emails.

DatabasesThis feature makes the robot possible to do the operation in the existing database.It’s important because database is the core resources for every company to do many data transactionThis function can be used in every database form such as MySQL, MsSQL, PostGres, etc.

Desktop automationThis feature makes us possible to automate tasks through the desktop applications.Most of operations in the office involving the desktop applications.We can use it for any desktop application such as Microsoft Office and etc.

Task recorderThis feature allows us to record every task that we run in the A.I. RPA designerBecause sometimes we need to re-use the same script for different operation.We can use this feature whenever we need to re-check or re-use the script.

Oracle appsApplication to manage the databases into the relational model in OracleBecause we may need to work with the Oracle app especially in every data transaction.We can use this function whenever we need to do operations with the Oracle app.

SAPSAP is an ERP solution.To work with SAPWe can use this function whenever we need to access the SAP system

CRM appsA system that used by the company to analyse all the data regarding the customer life cycle.It’s important because we need the output to create or improve better service for the customersCustomer service department.

ERP appsA modular software system designed to integrate the main functional areas of an organization’s business processes into a unified system.This is the core system in every organization’s business area.Since ERP is the core system, every department may use this system

Creating own workflowThe possibility to create our own workflow through the A.I. RPA designer.To provide the workflow so that the automation can be done.This is used in every automation processes.

Automate documentation
Easy way to capture user processFast documentation of the processes so that you can start to develop the scripts for the robotBy saving the business analyst time to document the processes, consultant can start to develop the scripts earlier.Business process re-engineering, details keystrokes for scripts.

 (only for Designer)
Mobile application clientThis an app in android phone or tablet to view and control the processes.You will be able to control the process from a phone. This is more user-friendly.More users with mobile phone.

Workflow forms & engine (by interacting with Control Tower)This will allow users to create their forms to integrate with Gleematic DesignerUsers can have a instantly feedback of the forms with GleematicCompanies who needs to collaborate with customers and suppliers etc.

(Using control tower)

(only for Designer)

*Both “Designer” and “Robot” have the functions (otherwise indicated). The difference is that users can make edits only in “Designer”.


Function in Gleematic

Gleematic Designer

Description: This “Designer” allows user to create, edit, modify scripts and steps in the scripts to automate tasks. It is typically used in development environments. The “Designer” also has the ability to create and execute scripts. There are 2 versions of Designer: the “Basic” and the “A.I.” version. See the section on “Functions” for more details.

Gleematic Robot​

Description: The “Robot” can be used in production environment to execute scripts (created by the “Designer”) to automate tasks. The “Robot” has the ability to store and execute scripts, which can be run one after another. There are 2 versions of Robot: the “Basic” and the “A.I.” version. See the section on “Functions” for more details.