Developers can spend 25% (or more) of their time finding and fixing bugs in their software. Part of this process involves trawling online forums and developer-focused platforms for answers, which can be a time-consuming endeavor — and is something Peritus.ai wants to help with.

The Palo Alto, California-based company has developed a machine learning-powered recommendation engine that surfaces the most relevant answers to all the questions people have about more than two dozen cloud-native technologies, including Kubernetes, Docker, Kafka, Elastic, MongoDB, Ngnix, Redis, and more.

The problem Peritus is looking to solve is that it can take a day or longer for the online community to respond to developers’ open source and cloud-native questions, and existing responses to similar questions are usually spread across multiple channels. Peritus is focused on bringing instant answers from trusted developer-focused sources, and has built a knowledge network of around 27 million questions, answers, and tips from online conversations funneled from Stack Overflow, GitHub, and numerous vendor forums.

The “Peritus recommendation engine” applies machine learning (ML) techniques such as named entity recognition (NER) to automatically identify key terms and then ranks all of the information to showcase the best answers.“We train our recommendation engine to understand the context of a long-form or narrative question specific to cloud-native topics,” Peritus cofounder and CEO Robin Purohit told VentureBeat. “Our recommendation scoring weights many variables, including people or content that has successfully answered similar questions. In addition, our ML techniques identify the core topic of the question.”

Founded back in 2017, Peritus previously served to automate various services and incident responses for datacenters before pivoting to its current business proposition. To help attract more developers to its platform, the company today announced it has raised $3.4 million in seed funding from First Ray Ventures and Benhamou Global Ventures, taking its total funding to $7.8 million since its inception.

Slack assistant

Initially available as a Chrome extension that launched in March, Peritus delivers suggestions for developers seeking help on around 50 public forums hosted on Stack Overflow, Discourse, Lithium (Khoros), and Salesforce. As of this week, Peritus also launched an assistant for Slack, which has emerged as a critical tool for developers, as it serves as a centralized forum for feedback and information-sharing. According to IDC (in a report commissioned by Slack, admittedly), developers spend 21% less time finding and fixing engineering-related bugs when using Slack.

“Our initial target is small to medium cloud-native teams that are using Slack to collaborate remotely,” Purohit explained. “The typical user is a cloud-native developer or DevOps engineer.”

Using the same recommendation engine as the Chrome extension, the Peritus Slack Assistant interprets the context of a long-form technical question and then throws up what it deems to be the best answers, either via a direct message or through a reply to the developer, who can then decide whether to post it publicly to the channel.

Additionally, the Peritus Assistant for Slack recommends three top experts from across the online sphere that developers can contact for more specific support.

The Peritus Assistant for Slack is now available for free to anyone who signs into the Peritus Slack Sandbox. But for teams focused on collaboration, the Peritus Assistant for Slack can be downloaded directly to a Slack workspace starting at $10 per user per month, which includes a 30-day free trial. For companies that use Slack as a community or support channel, an annual subscription is available upon request, which includes various intelligence analytics and an unlimited number of users.

But perhaps more importantly, companies on this premium plan can also configure the assistant to ingest their own Slack conversation archives to improve the Peritus recommendation engine. By default, these conversations become part of Peritus’ public dataset to improve its ML models for all Peritus users. “This solves a critical limitation for Slack communities, since the conversations are not available on Google Search,” Purohit explained.

However, Peritus also provides an option to keep recommendations from a team’s Slack conversations private, making them available only to the members of that group.

Developer-driven

While providing timely, accurate responses to some of the most urgent cloud-related questions is Peritus’s bread and butter, the Chrome extension also allows users to answer questions and earn “reputation points.” This can help community managers, for example, establish who the most active and helpful contributors are, or what forums are the most engaged with a particular technology or company.

Above: Peritus provides an analysis of community members to “identify and nurture more experts.”

Developer engagement has become a core focus for companies of all sizes, given that developers increasingly drive buying decisions in companies. And engaging developers can also be a lucrative mechanism for attracting top technical talent.

All of this is why Microsoft shelled out more than $7 billion to buy GitHub. It’s also why Peritus focuses heavily (though not exclusively) on “open” rather than proprietary technologies, as these communities are well-aligned with the problem that Peritus is trying to solve.

“Our focus is open source and open core technologies, where there is a strong culture of knowledge sharing,” Purohit explained. “Our belief is all companies, even those who have traditionally had closed business models, must rapidly move to this developer-first approach by opening APIs and PaaS (platform-as-a-service) layers and building open technical communities.”

In the immediate future, Purohit said his company plans to extend its Chrome extension to support developers asking questions on GitHub, Reddit, StackExchange, and Stack Overflow’s business-focused Teams, though the goal is to eventually go much wider than that.

“Our future plans include enabling the Chrome extension for any web page,” he said.

Effectively, this means a developer could be asking a question on just about any public website, from Hacker News to Facebook, and receive instant answers to their questions.

“Our mission at Peritus is to be the ‘developer success’ company, providing the millions of developers powering the software economy with the expert knowledge they need to be productive with the latest open source and cloud technologies,” Purohit added.

‍Originally published via VentureBeat

‍https://venturebeat.com/2021/09/24/peritus-ai-automates-answers-for-cloud-native-questions/