Devin 1.2: An updated AI engine enhances text with intelligent internal logic, including voice

Last year, Information introduced the AI ​​feature with a feature called Devin – the first of the world AI engineer. The offer was closed for a few months, but now it is available and learn new information. Case in point: The Scott Wu-led startup has just released Devin 1.2, which brings more power to take the AI ​​engineer’s ability to handle all development tasks to a whole new level.

The highlight of Devin 1.2 is the improved internal logic, which makes the developer better at handling and reusing code. It also includes the ability to take voice messages through Slack, giving users a seamless way to tell Devin what to do.

The development comes at a time when AI assistants are being touted as the future of technology. Experts believe that in the near future there will be a time when people and assistants will work together, who used to do repetitive tasks (which has already started to happen). Recently, at CES, the boss of Nvidia Jensen Huang he said that in the future, IT departments of companies will transform into “HR departments” of AI, responsible for deploying and managing assistants who perform various functions in the company.

What does Devin 1.2 bring to the table?

Although it is not a major change, Devin 1.2 brings some interesting features to make the agent better at his job. The first step here is to be able to think in depth in the storage. This means that Devin can now better understand the structure and content of the database.

With this understanding, the agent can identify which file is related to a specific task, identify and reuse existing codes and processes, and be more accurate in recommending changes or creating pull requests (PRs), reducing errors and manual changes.

For developers, this capability can mean speeding up work and reducing knowledge from searching files, understanding codebases or fixing inconsistent code.

Another notable change with Devin 1.2 is the introduction of voice messages. Devin can also take voice messages from users, via Slack.

All you have to do is put Devin in a Slack chat, click the “Record audio” button and describe the tasks or solutions the AI ​​engineer needs to do. Devin is ready to take action step by step and start commanding using his development tools – his shell, code editor and browser.

This move streamlines the way a person interacts with the agent, saving the trouble of typing natural language into Devin’s chatbot interface.

Inbound management, business innovation management

Awareness has also led to changes in Devin’s performance.

For example, in the new release the company is introducing machine previews to simplify the way to access Devin’s workspace.

“When you enter Devin while chatting with Devin’s browser, we will save a cookie for future sessions (if the cookie expires, you will need to provide Devin’s information again in Privacy). This also opens the authentication process that is required to visit a link on Devin’s machine,” the company wrote in the documents. blog post.

Cognition is also introducing business accounts, where organization admins will get a central interface to manage multiple Devin workspaces, including members and their logins, and billing them.

Finally, the company is adding a pay-as-you-go option, allowing users to pay additional fees beyond their subscription limits. This way, when users finish allocating ACUs each month, they can continue to build beyond that limit by paying extra.

The model has been active since January 9, where users can set their own additional budgets according to their needs. This allows users to control their spending and ensure that the service is not interrupted when they need extra power.

Currently, Devin is available to support engineers at a starting price of $500 per month – with no seat limit. A number of businesses are already integrating into their channels, including Lumos, OpenSea, Curai Health, Nu Bank and Ramp.

Devin’s new capabilities come as competition in the AI ​​engineering space heats up. From GitHub Copilot based on the spread of Magic and Poolside AI to raise more money to create a better skill, a competition to create the final product. AI code assistant it is growing. Each player is striving to redefine software development, promising faster workflows, less information, and seamless collaboration between humans and AI.

As these AI-powered assistants continue to evolve, they are not only changing the way designers work but are shaping the future of modern work, where efficiency and innovation are driven by the collaboration between humans and machines.

By 2028, Gartner estimates, 33% of company software includes AI, enabling autonomous decision-making in 15% of daily tasks.