Brevian is a no-code enterprise platform for building AI agents

6 Min Read

Sunnyvale-based Brevian desires to make it simpler for enterprise customers to construct customized AI brokers. At the moment, the corporate is concentrated on help groups and safety analysts — areas the place each the use instances and coaching units are effectively outlined — however plans to increase to different areas over time. The corporate is popping out of stealth as we speak and saying a $9 million seed funding spherical.

Brevian, which types its identify as BREV/ΛN, was based by Vinay Wagh (CEO) and Ram Swaminathan (CTO), who got here to this startup journey from very completely different instructions. Wagh was beforehand a product director at Databricks and the pinnacle of merchandise at Bracket Computing. Throughout his time at Databricks, he labored on getting the corporate able to promote to massive enterprises after which launch its serverless product, Databricks SQL Serverless. “That went GA in February, and I used to be all the pieces happening, Wagh mentioned. “Once I noticed ChatGPT and different issues, for me, it was simply: how is that this relevant within the enterprise? Immediately, our alignment was that every one these lacking items are going to be wanted. There’s a lot promise, however except any person solves this piece, it’s not going to work effectively. That was the inspiration to get on the market — there’s stuff occurring. After which I met Ram [Swaminathan] by means of a standard buddy.”

Whereas Wagh labored loads on the information aspect, Swaminathan’s background is on the pc science and machine studying aspect. After working in academia for some time, he spent 9 years at Bell Labs and seventeen years at HP Labs, engaged on all the pieces from coding concept to cryptography and early machine studying efforts. Then he went to LinkedIn, the place he headed up the AI belief crew earlier than leaving in late 2022 to affix Wagh in founding Brevian as its CTO.

Brevian founders Vinay Wagh (l.) and Ram Swaminathan (r.).

Brevian founders Vinay Wagh (l.) and Ram Swaminathan (r.).

“I didn’t actually have a product sense,” he informed me. “You write papers, you write patents — however can you actually construct a product that may influence a lot of folks? And LinkedIn was an ideal goal: go there, you assist shield a platform that serves jobs and connections to folks. That’s it. It’s a fantastic influence on society. I actually get pleasure from working with product folks and engineering, and coordinate all this stuff and actually create worth.”

See also  Why Healthcare Leaders are Seeking Streamlined AI Adoption and Other Takeaways From the Mayo Clinic Platform Conference - Healthcare AI

On the core of all enterprise initiatives round generative AI is safety, the crew argues. AI distributors want to make sure that no private info leaks, for instance. When the corporate launched, lots of enterprises determined to ban ChatGPT inside their organizations as a result of there was no technique to assure that workers wouldn’t feed delicate knowledge into the chat prompts.

In its early days, Brevian focussed totally on safety. “We rapidly circled fashions that detect [personally identifiable information],” Wagh defined. “We seen that this immediate injection assault vector was rising, so we constructed an intent-based system to detect immediate injection. And so what we ended up doing at the moment […] was amassing all of the prompts that break alignment, and we ran a mannequin throughout all of them and it was catching all of them.”

Picture Credit: Brevian

Apparently, Wagh famous, that even as we speak enterprises fear loads about safety within the context of enormous language fashions, however they don’t truly fairly know what they’re actually fearful about past these leaks. The actual problem, the crew rapidly realized, was that what was holding enterprise adoption again wasn’t truly a lot safety however the problem of constructing techniques that solved actual issues for them.

“Our imaginative and prescient was to allow enterprise customers within the enterprise to have the ability to use AI to simplify their each day duties. All of that slot in throughout the imaginative and prescient to increase from safety and construct these AI brokers.”

Brevian’s seed spherical was led by Felicis companion Jake Storm (and this marks Felics’ largest seed spherical test but). He famous {that a} yr in the past, everyone was speaking about AI infrastructure, however few folks had been speaking about purposes — and the significance of securing these. “2023 was form of [the year of ]AI infra. 2024 is the yr of AI apps — and we simply felt like they had been to this point forward,” he informed me.

See also  Building a viable pricing model for generative AI features could be challenging

With this funding spherical, Brevian goals to speed up product improvement and increase its crew to fulfill rising buyer demand by means of its early launch program.


Source link

Share This Article
Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *