With $9 million in funding, Brevian Inc., a firm that offers artificial intelligence agents for jobs like looking into data breaches, officially began today.
Felicis Ventures led the investment, according to TechCrunch. The transaction is purportedly the largest seed investment made by the venture capital firm to date.
Because large language models are configurable, users can modify them to suit jobs that the models may not be well-suited for right out of the box. An organization could, for instance, provide an LLM with access to a technical support knowledge library so that it can handle help desk tickets. These kinds of customized models are commonly called agents.
Brevian, a company located in Sunnyvale, California, provides a platform that lets employees construct AI agents without having to write any code. At the moment, the platform primarily concentrates on automating manual processes in two domains: customer support and cybersecurity.
Administrators frequently need to examine logs from various systems in order to determine whether unexpected network behavior could be the consequence of a breach. Investigations are slowed down when those logs must be manually retrieved and examined, which can take a long time. Brevian claims that its platform enables the development of AI agents that can automatically obtain information about probable security vulnerabilities.
Teams working on cybersecurity can communicate with the agents using an interface akin to ChatGPT. For instance, an administrator could instruct an AI driven by Brevian to obtain data regarding recent connections to other networks. Additionally, you can filter the data: It is possible to direct the agent to show only connections that have exchanged more than 500 megabytes of data in the recent 24 hours.
Large volumes of information must also be reviewed by customer support staff on a daily basis. When troubleshooting a server issue, for instance, a help desk staff might need to look through several technical guides and logs from the data center in question. AI agents that can autonomously retrieve customer support information can be created thanks to Brevian.
The company says that their platform has a “reasoning engine” that enables it to do intricate analysis in addition to retrieving data. The underlying AI algorithms are prevented from producing dangerous responses by an integrated moderation mechanism. Furthermore, the way in which responses are generated makes it reasonably easy for users to verify their accuracy.
Brevian has added cybersecurity safeguards to its platform. Administrators have the ability to control who among employees has access to a certain AI agent and how. Brevian comes with an audit tool that records user interactions with agents as an extra precaution.
According to reports, the company plans to expand its workforce by using the money from its recent seed round. Brevian intends to step up its efforts in product development concurrently. The startup plans to add additional use cases to its platform, moving beyond cybersecurity and customer assistance.