Email&Tasks: Jira Cloud for Gmail is a collection of three add-ons: Atlassian Marketplace, Google Workspace Marketplace connector for Gmail, and a Chrome browser extension. Its primary purpose is to enhance task management processes by giving users easy access to tasks and various related operations. Additionally, the add-on can identify tasks within emails in a user’s inbox. We have provided a more detailed explanation of how the add-on functions in our article – How to increase your productivity with Jira and Gmail integration .
At the heart of the Email&Tasks app by SaaSJet is its ability to convert emails into tasks in Jira. This can be done manually or with the revolutionary 1-CLICK TASK option, which utilizes an AI solution. The team dedicated significant effort to enhancing the app’s automatic task detection capabilities and integrated GPT-3 technology into the app. This article will discuss the reasons behind this decision, the preceding steps, and the outcomes achieved as a result.
Introduction to GPT-3
OpenAI developed the Generative Pre-trained Transformer 3 (GPT-3), a natural language processing model based on a deep-learning architecture called a transformer. With 175 billion parameters, GPT-3 is the power version of the GPT series and has a range of capabilities, including translation, question answering, completion, and generation. GPT-3 can comprehend natural language and produce human-like responses, making it useful for chatbots, virtual assistants, and speech applications. Additionally, GPT-3 can be customized for specific tasks, such as sentiment analysis, generalization, and language modelling.
A notable feature of GPT-3 is its ability to perform multi-shot and zero-shot learning. Multi-shot learning involves training the model with only a few task examples. In contrast, zero-shot learning enables the model to perform tasks it has not been explicitly trained for by utilizing existing knowledge and context.
Integration of GPT-3 into Email&Tasks:Jira Cloud for Gmail app
Our team previously utilized the Codeq API to identify tasks within the email content. Codeq’s NLP technology, which employs sophisticated deep learning models and linguistic analysis tools, enables the extraction of detailed representations from text data. The Codeq API is designed to provide high levels of accuracy and precision when it comes to an understanding and categorizing tasks. However, despite its impressive capabilities, we found it somewhat lacking and ultimately opted to switch to GPT-3 for this task.
What is the difference? Let’s try to compare these two technologies.
|Purpose||Natural language is recognized, but the result is often negative in terms of quality||Wide range of natural language processing tasks|
|Accuracy||High||Very high accuracy in natural language recognition. Much better than Codeq API|
|Model architecture||Deep learning||Transformer-based architecture|
|Performance||Highly specialized for task classification||Broad-based but may require fine-tuning for specific tasks|
|Language Support||English||Supports a wide range of natural languages|
|API availability||Available for commercial use||Available for commercial use|
|Cost||Paid subscription||Paid subscription|
|Training data||Labeled text||Large-scale general-purpose text corpora|
Since the application deals with natural language, it made perfect sense to transition to this technology. To illustrate how email detection works using GPT-3, let’s consider an example where the technology has already been implemented.
The Applications and Advantages of GPT-3 in the Email&Tasks: Jira Cloud for Gmail
How does it work?
- First, you need to install the app and its connector. You can find them in the Atlassian Marketplace and Google Workspace Marketplace, respectively. Once you’ve done that, go to the email you want to turn into a task and click the Email&Tasks icon.
2. Select the TRY 1-CLICK TASK (AI-BASED) option.
3. Once the app identifies your email content, it generates a draft for the next task, taking into account the priority mentioned in the text and assigning the recipient as the Assignee, all without any further action required from you.
Now you just need to click the Create issue button.
By simply taking a few seconds, you can complete the task effortlessly. If the email comprises attachments, selecting the “Include Email Attachments” option will automatically attach them to the issue.
Aside from integrating GPT-3 in the Email&Tasks: Jira Cloud for Gmail app, we have other plans to enhance its performance. Our goal is to improve user experience and streamline workflows. We encourage you to try our AI-powered app and share your feedback. You can avail of our 30-day trial to experience its advantages. Additionally, please note that the add-on is free for teams with up to 10 users. Enjoy!