Home AI News Pienso builds no-code tools for training AI models

Pienso builds no-code tools for training AI models

0
Pienso builds no-code tools for training AI models

AI could be the “it factor” of the second. However that doesn’t imply it’s getting simpler to deploy.

In line with 2023 S&P International survey, about half of corporations with a minimum of one AI undertaking in manufacturing are nonetheless on the pilot or proof-of-concept levels. The explanations for the sluggish ramp-ups range, however the generally cited ones are challenges round knowledge administration, safety and compute sources.

Of companies responding to the S&P ballot, round half stated they aren’t able to implement AI — and gained’t be for 5 years or extra.

Fortuitously, there’s an rising variety of merchandise from startups and Huge Tech distributors alike that goal to deal with these AI deployment roadblocks. (See Google’s ML Hub, Kore.ai and Viso to call a number of.) One of many newer entrants is Pienso, a platform that lets customers construct and deploy fashions with out having to put in writing code.

Birago Jones and Karthik Dinakar based Pienso in 2016 primarily based on their analysis at MIT (they’re alumni). The 2 met a number of years again at MIT’s Media Lab as graduate college students. 

“We teamed up for a category undertaking to construct a device that might assist social media platforms average and flag bullying content material,” Jones, who serves as Pienso’s CEO, informed Trendster in an interview. “There was only one downside: whereas the mannequin itself labored the best way it was speculated to, it wasn’t educated on the appropriate knowledge, so it wasn’t in a position to establish dangerous content material that used teenage slang.”

Jones and Dinakar ultimately realized that the answer was to have subject-matter specialists — on this case, youngsters — assist prepare the mannequin. They constructed instruments for this goal, and a few years later, Jones and Dinakar teamed as much as commercialize these instruments.

What resulted was Pienso, which Jones describes as an AI suite constructed for “non-technical expertise” — specifically researchers, entrepreneurs and buyer assist groups who’ve entry to massive quantities of information for AI coaching, however lack the mandatory sources to construction and analyze it.

“A lot of the AI dialog has been dominated by … massive language fashions,” Jones stated, “however the actuality is that nobody mannequin can do every little thing. To have the ability to scale to AI’s full potential, the place it could actually handle enterprise processes and work together with clients, you’ve got to have the ability to prepare and fine-tune your mannequin. Pienso believes that any area professional, not simply an AI engineer, ought to be capable of do exactly that.”

Pienso guides customers by means of the method of annotating or labeling coaching knowledge for pre-tuned open supply or customized AI fashions. (It is dependent upon the mannequin, however AI usually wants labels — like a picture of a hen paired with the label “finch” — to study to carry out a process.) The platform, which could be deployed within the cloud or on-premises, integrates with enterprise techniques by means of APIs. However it could actually additionally function with out APIs or third-party companies, holding knowledge inside a safe setting.

Sky, the U.Okay. broadcaster, is utilizing Pienso to investigate customer support calls, Jones says, whereas an unnamed U.S. authorities company has examined it to observe unlawful weapons monitoring.

“Pienso’s versatile, no-code interface permits groups to coach fashions instantly utilizing their very own firm’s knowledge,” Jones stated. “This alleviates the privateness considerations of utilizing … fashions, and in addition is extra correct, capturing the nuances of every particular person firm.”

Corporations pay Pienso a yearly license primarily based on the variety of AI fashions they deploy. The larger the variety of fashions, the upper the licensing value.

“We deliberately designed our pricing to permit clients to check out fashions beforehand to grasp how AI might help them with out first having to make a large funding,” Jones added. “We wished to provide clients the liberty to experiment with constructing new fashions earlier than deploying them.”

It’s a enterprise mannequin that appeals to traders, apparently. Pienso not too long ago raised $10 million in a Sequence A funding spherical led by Latimer Ventures with participation from Gideon Capital, SRI, Uncork and Good Development Capital.

Bringing Pienso’s complete raised to $17 million, Jones says that the brand new money might be put towards scaling up Pienso’s gross sales, advertising and marketing and buyer success groups, recruiting engineering expertise and constructing new options for the platform.

Stated Latimer Ventures’ Luke Cooper in a press release: “We always hear about the necessity to democratize AI, however what makes Pienso stand out is how they’re interested by the position of a site professional on this equation. They’re empowering those that perceive their knowledge greatest to get probably the most insights out of it. It’s fostering a future the place we’re constructing smarter AI fashions for a selected utility, by the people who find themselves most accustomed to the issues they’re making an attempt to unravel.”