Okta on Thursday agreed to acquire AI identity security startup Permiso Security, betting that demand for protecting AI agents and other machine identities will grow as enterprises deploy autonomous software across their operations.

The identity management company did not disclose the terms of the transaction. But TechCrunch has learned that the acquisition is valued at just under $200 million and is structured as an almost all-cash deal, according to a source with knowledge of the deal. A spokesperson for Okta did not dispute the figure when asked by TechCrunch, but would not comment on specifics of the deal terms.

The deal is expected to close in the third quarter of its fiscal 2027, Okta said, subject to customary closing conditions.

Okta’s move to buy Permiso comes as identity management companies seek to expand beyond verifying users at login to continuously monitoring what users, applications, and AI agents do once gaining authorized access to a network environment. That shift has intensified competition to secure machine identities as enterprises embed AI deeper into everyday operations.

Permiso, which emerged from stealth in 2022, develops software that helps security teams spot suspicious activity in cloud environments after users or applications have been granted access. More recently, the startup has expanded its platform to monitor AI agents and other machine identities.

Co-founded by former FireEye executives Paul Nguyen and Jason Martin, Permiso specializes in detecting attacks that use stolen or compromised identities to move through cloud infrastructure. In April, the startup also introduced SandyClaw, a platform designed to analyze AI agent skills in a sandboxed environment to identify malicious behavior before they are deployed.

The deal strengthens Okta’s push into securing AI agents and other non-human identities alongside its core identity management business.

“Permiso will extend Okta’s identity security fabric with proven identity threat detection and response capabilities, and an incredible threat research and security team that will advance Okta’s threat detection and prevention capabilities,” Okta’s chief product officer Ely Kahn said in a prepared statement.

Permiso has raised about $29 million to date, including an $18.5 million Series A round in April 2024 led by Altimeter Capital. People familiar with the financing said the Series A valued the Palo Alto-based startup at about $80 million on a post-money basis.

Rumor mill: Google’s Pixel 11 is coming into focus thanks to two separate leaks. One is an official teaser video released by Google itself, while the other is a Bluesky post from WinFuture’s Roland Quandt detailing the phone’s storage options, battery capacity, and pricing.

The Google video is the more visual of the two. It offers a close look at the Pixel 11 Pro’s rear camera bar, where the phone retains its three-lens setup while adding a circular light effect beside the cameras. The clip does not name the feature or explain what it does, but the spinning colors suggest Google is building something new around the camera area rather than adding a feature elsewhere on the phone.

Google’s narration only mentions familiar services such as Gmail, Maps, Search, Calendar, and Gemini. The video title teases “the next obvious move,” but Google leaves the meaning deliberately vague. That has fueled speculation that the light could serve as a status indicator, a Gemini-related processing light, or a notification system, but none of those theories have been confirmed.

The second leak comes from Bluesky, where Quandt posted what appears to be retailer-sourced information about the Pixel 11 lineup. According to the post, 256GB will be the minimum storage option across the entire range, marking a long-awaited shift for Google’s standard models. The leak also claims that the Pixel 11 will cost £879 in the UK, while the Pixel 11 Pro and Pixel 11 Pro XL will be priced at £1,079 and £1,279, respectively.

Those pricing figures are not as straightforward as they initially appear. The Pixel 11 and Pixel 11 Pro are technically cheaper than their 256GB Pixel 10 counterparts, while the Pixel 11 Pro XL costs more than last year’s model. However, the difference stems from Google’s decision to eliminate lower-capacity storage tiers rather than from a broad price cut across the lineup.

Pixel 11 256GB = 879 GBP, 4985mAh Pixel 11 Pro 256 GB = 1079 GBP, 4850mAh Pixel 11 Pro XL 257 GB = 1279 GBP, 5115mAh

– Roland Quandt (@rquandt.bsky.social) July 28, 2026 at 3:06 AM

Quandt’s Bluesky post also includes battery capacity details. The base Pixel 11 is said to feature a 4,985mAh battery, while the Pixel 11 Pro and Pixel 11 Pro XL are listed with 4,850mAh and 5,115mAh batteries, respectively. If accurate, the base model and the larger Pro XL would receive modest battery upgrades, while the regular Pro would see a slight reduction compared with the previous generation.

Google is set to unveil the Pixel 11 lineup on August 12. Until then, the picture remains incomplete.

Imagine you’re a company. You’ve got sensitive data you need kept safe, and you’ve got a cloud service provider that’s protecting it. But do you know how far that protection goes, and at what point it’s your turn to take over?

The sheer number of data breaches due to human error, specifically database misconfiguration, suggests that many don’t know that they, not the provider, are responsible for securing all the data properly (or are unsure how to do it).

from the needlessly-stupid dept

The American healthcare system and the portion of the government in charge of our health have been a failure of record proportions for the second consecutive year. In 2025, America set a record for collecting the most confirmed cases of measles since 1991, with 2,289 confirmed cases. This year, America already blew past last years case total, where we now sit at 2,318 confirmed cases. We’re only about 58% of the way through the year, in case that little fact escaped you, which means we’re very likely to have many, many more total cases this year than last.

That is happening despite healthcare professionals and some media members sounding the alarm bells over all of this going all the way back to February of 2025! If you need the government’s perspective on this embarrassing failure, one in which we’ve allowed a disease officially declared eradicated to roar back into existence all because there are some liars out there who scare people about vaccines, well, RFK Jr. thinks he’s doing a bang up job managing the whole thing.

He hasn’t. He’s a large part of the reason this now 19 month outbreak exists to begin with. He hasn’t mobilized any vaccination efforts, preferring to leave that to the states. His messaging around vaccination and how to handle the measles has been so all over the place that he might as well have no stance on the matter at all. He’s checked out from most of his job, by some reports.

And as all of this failure is staring us in the face, Kennedy is still allowing the anti-vaxxer messaging to spread undisputed and uninterrupted, which is the precise reason the measles has returned in the first place. And it’s very likely to get worse.

“There were two healthy little girls in West Texas that died from measles,” said Dr. Paul Offit, Director of the Vaccine Education Center at the Children’s Hospital of Philadelphia, in an email to TIME. “Those were the first measles deaths in a child in this country in more than 20 years. I think it is fair to say that measles is epidemic in the United States. And if people are coming from outside this country to the United States, they better make sure they’re up-to-date on their immunizations.”

Adds Dr. Peter Hotez, Dean of the National School of Tropical Medicine at the Baylor College of Medicine: “For all practical purposes measles has returned to the United States, both in terms of cases and hospitalizations, as well as two needless deaths.  We should expect our annual winter-spring surges of measles to get even worse in subsequent years.  And it won’t be just measles—pertussis is also returning, and other childhood infectious diseases might follow.”

Measles has this fun little curve ball it likes to throw at its victims: immune amnesia. One of the more insidious things this virus does is destroy T-cells and B-cells, which are the long-term immunological memory cells. They’re the part of our bodies that recall infections from long ago and build in the longer lasting immunity we get either from infections themselves or from vaccinations. Allowing measles to get worse means not only dealing with that disease, but potentially many others that could come roaring back. Like pertussis. Or polio.

We haven’t developed therapies for measles in decades. We didn’t have to. Vaccination rates meant there was nearly no disease for which therapies were needed. That appears to be changing now, with reports that researchers are starting to develop drug therapies due to the record case counts.

That is valuable time, expertise, and money that is being spent to treat a disease that we were fucking rid of two decades ago. Thanks to liars and charlatans like RFK Jr., combined with a percentage of our populace that is uncaring of its own community.

I very much do not look forward to writing about breaking this record again in 2027, but I fear I will be.

Filed Under: anti-vaxxers, epidemic, health & human services, measles, rfk jr., vaccinations

Sunreef Yachts and BRABUS just put their names on the same boat for the first time. The result is the ULTIMA 55, a 55.8-foot carbon-heavy catamaran that treats open water the way BRABUS treats asphalt. Two 10.8-liter six-cylinder engines sit in the hulls, each making 533 kW or 725 horsepower. Combined output hits 1,450 horsepower and the boat clears 40 knots. An advanced hydrofoil system lifts the hulls higher as speed builds, cutting drag and almost erasing heel so the ride stays level and surprisingly quiet even when the throttles are pinned.

The length comes in at 55.8 feet, the beam is 18.4 feet, but when the lateral folding platforms are dropped, that increases to 24.9 feet, and you have an open-ended terrace at the back, similar to a private beach club rather than a normal yacht’s cockpit. Then there’s the Esthec decking, which provides a sturdy footing whether wet or dry. The transformer sunpads change from comfortable loungers to upright seating options. They have an extensible carbon fiber dining table, an electric BBQ, and a wet bar that are ready for anything from a quiet afternoon to a big party. Up ahead, there is a separate central sunpad where you may spend some peaceful time away from the action.

LEGO Icons Captain Jack Sparrow’s Pirate Ship Building Set for Adults, Ages 18+ – Room Decor for…

• BUILDING SET FOR ADULTS – Embark on a creative voyage with the LEGO Icons Captain Jack Sparrow’s Pirate Ship (10365) building set for adults
• 9 LEGO MINIFIGURES – Pose Jack Sparrow, Will Turner, Gibbs, Cotton, Anamaria, Marty, Elizabeth Swann and Hector Barbossa minifigures around The…
• AUTHENTIC DETAILS & FEATURES – Rotate dials to deploy the port or starboard cannons, turn the ship’s wheel to steer the rudder, place minifigures in…

Carbon fiber is now popping up in all the right places. So you’ve got exposed front and back wings, pillars with illuminated BRABUS emblems, side ventilation elements, and you know what truly caps it off? The unique roof rack features carbon-fiber LED lights, which serve as both functional lighting and a signature design element. Owners can choose between Graphite Black and Phantom Grey for the hull, as well as BRABUS Graphite, Red, or Pearl exterior upholstery.

Stepping inside the cabin, you’ll feel as if you’re in a car, since the BRABUS automotive interiors have been directly adapted into the yacht. Moonstone Masterpiece leather covers the seating, with the signature Triangle stitching pattern. The area is finished with Alcantara headlining, exposed carbon fiber, aluminum accents, and unique ambient lighting. You’ve fully outfitted the kitchen, with a 43-inch TV that pops up from the counter when needed and, of course, Bowers & Wilkins Marine Series speakers to handle the music. The nicest feature are the height-adjustable tables, which change the saloon into a dining or lounge area in an instant.

The helm area follows the same principle, with a single widescreen BRABUS interface that brings all navigation and system management together in one spot. The engine controls are right there for when you need them, and the joystick operation allows the boat to spin on a sixpence for snug docking.

You can select between two cabin configurations, with the regular one featuring a master suite, a VIP stateroom for four guests, and a crew cabin. The other option includes a third cabin for six guests and one crew member to remain overnight. It’s all the same down there, with wood veneers and more carbon fiber. So the lower decks remain consistent with the rest of the boat. The specifications include 1,500 liters of fuel, which is ample given that you have 330 litres of freshwater on tap, as well as a 10.5 kW generator to meet your electrical demands. The maximum draft is 4.6 feet, and the boat is CE Category B approved for 12 passengers or C certified for 20.

Sunreef handled the most of the naval design, then turned it over to BRABUS to give it a visual identity, add some high performance with their materials, and put their visual stamp on the project. The ULTIMA 55 is the first product of that relationship, as well as the first to be delivered to Sunreef’s Ras Al Khaimah factory in this configuration. In terms of price, the basic non-BRABUS Ultima 55 has been advertised (before any additional fees) at around the mid-two million mark, and the BRABUS version is almost certainly more expensive.

As X-Men ’97 continues the epic saga of Marvel’s mutant heroes, voice actor Jennifer Hale returns as one of the team’s most iconic members, Jean Grey. After the events of season 1, the show follows the X-Men, who have been scattered across history, as they try to return to their original era and prevent the rise of their most powerful enemy: Apocalypse.

In an interview with Digital Trends, Hale discussed returning to voice Jean Grey after doing so in many of Marvel’s animated projects. She also broke down the evolution of Jean’s character and what fans can expect from season 2 of X-Men ’97.

This transcript has been edited for length and clarity.

Digital Trends: How does it feel to be voicing such an iconic character in such a popular show right now? 

Hale: Man, I feel like I… I won the golden ticket. It is incredible. I love it. This show, [when] I first booked this, I was thrilled.

I love the director. I have to shout out our voice director, Meredith Layne. She’s incredible. Voice directors are these sort of unsung heroes that make us all sound incredible. And she really, especially with the role, what the team has done with Jean in this incarnation of the show. 

There’s Jean, there’s Goblin Queen, there’s Madelyne Pryor. There’s all these iterations, and to have Meredith there as a guide is incredible. And to be part of this cast and to follow in Catherine [Disher]’s footsteps is just such a, such a huge honor. And I love this show. 

I love this team. This team is insane. The writing is incredible. And what I love is that so many of the production crew grew up watching the show. You can see it, right?

Digital Trends: Yeah, the passion is there, and I’m a huge fan of the show as well, and you did a fantastic job playing Jean Grey. This isn’t actually the first time you voiced her. You portrayed her in previous Marvel animated shows. Why do you keep returning to the character after all this time?

Hale: You can’t stop me…the thing about the voiceover world, the actors, my peers in the voiceover industry are ridiculously talented. And a few of us get to be Jean, and I am so grateful that I get to be her in the places that I am. I absolutely love it. I could go on about it, obviously.

Digital Trends: Oh, well, please do. 

Hale: Yeah, it’s amazing. It’s just absolutely amazing. It’s so funny.  This whole thing around sharing roles…I’ve been Rogue before, and Lenore [Zann] is just incredible as Rogue. She’s the OG, right? 

Yeah. And she’s amazing, and she’s just such an incredible human being, too. Yeah, this whole cast, I mean, I don’t even want to start, ’cause I’m gonna leave people out, and then I’m gonna feel terrible, but every single member of this cast, like, ‘Hello, what a great job.’ And our newer people. Oh, my gosh, the new additions this season. Look out!

Digital Trends: Yeah. I know you got the younger Cable in season 2 from Obsession actor Michael Johnston. He did a really good job in that role. Honestly, I couldn’t believe it was him. And speaking of which, like a big part of the show’s season was Jean and Scott’s relationship with their son, Nathan. It expands on what happened in season one. Can you tell us what makes it so special in season two?

Hale: Oh, the fact that we actually have to see him again…we gave him up in season 1. We didn’t know…we just had to trust, and it didn’t go the way we planned. That’s all I’m gonna say for those who don’t watch. But to get to see him again, to get to be in his life again, is just, ‘Wow, what a great, great moment!’ Right? 

Digital Trends: Right. And so what has been your favorite moment in the series so far?

Hale: All of it? I don’t know. I love…Oh, man. That’s really hard to say. Don’t make me pick. That’s what I always say. Yeah.  I think I just, I have to say, what you just touched on, the reconnecting with Cable. That whole arc. I just love that so much. And I’m not gonna say anything else, ’cause I don’t want to accidentally hit spoilers.

Digital Trends: Right. And without giving away any spoilers, what else can fans expect from this season of X-Men ’97? 

Hale: Mm, that is the tightrope. Don’t know that I can walk [it]. So all I’m gonna say is buckle up. All right. 

Digital Trends: I’m buckled up. And now, tell us, what is your approach to bringing these iconic characters to life as a voice actor? You’ve done so many different roles, and I bet it’s just a different thing each time. How do you do it? 

Hale: It’s all about the writing. Writers make the world go round. It all starts with the writers…and these are incredible scripts. Everything’s there. So it really begins and ends with the writing.

I bring in what my creativity says, and then there’s the voice director, and the voice director’s taking everything that the whole production team is saying, and funneling it through her to me, or to us actors. And so, that is incredibly important. And staying open with this show is very interesting. 

When we did season one…the first time we recorded it, we recorded it in this total OG style, with the ’90s acting and the whole thing. So it was very different….and so we went back and tinkered with it, and we went away from that, and then back to it, and then we kind of found this middle ground where we get to honor it.

And most of the time, the way we’re set up, we’re recording solo, but there are times when Ray [Chase] and I get to record together, and Ray is so incredible. There have been moments when we…we actually got to be in the actual booth together at the same time, and that was amazing. But we can also do remote together, and just playing off each other is so valuable, so incredible.

Digital Trends: Absolutely. And… you’re already working on seasons 3 and 4 of the show. How’s that going so far? 

Hale: Oh my gosh. It’s incredible. I can’t wait for you guys to see it. I just can’t wait. I’m backing up to your earlier question about the moments that I love. I was just thinking about some of this stuff with Morph and Polaris and Emma Frost…everybody’s getting a moment, and I love that. Absolutely love that.

Digital Trends: Me, too. I mean, considering you’re already working on the fourth season of the show, how long do you think it could go on?

Hale: Well, we’re at 30 years now, so who knows?

Digital Trends: On that note, like what other projects do you have in store for fans?

Hale: I can always talk about Unicorn Academy, ’cause that’s going on right now, and I know it’s safe to talk about. We just wrapped the game called The Long Dark, which just wrapped our last chapter, and it’s incredibly beautiful. I’m working on some game stuff that is insane and some very cool animated stuff coming out. But again, NDA. Can’t talk about any of it…and then, I’m working on SkillsHub all the time, my site for actors.

Digital Trends: Fantastic. Is there anything else you’d like to say to the fans?

Hale: I want to say a huge thank you to the fans. Thank you, Anthony, for this time. And thank you, Digital Trends, for having me on. The fans, oh my gosh. You guys are our community, because you’re there, we get to do this, and we do this for you. So we are totally in symbiosis, I think it’s called. We are totally a circle. So just, thank you so much.

X-Men ’97 is now available to stream on Disney+.

Executive search firm Christian & Timbers estimates that there are only about 2,000 engineers in the U.S. with the special cocktail of sector know-how, gravitas, and hands-on applied AI experience needed to consistently help enterprises see a return on their AI expenditures. 

“Not 2,000 available,” reads the study, shared exclusively with TechCrunch. “2,000 total.”

As enterprises move from trying to access the best models to figuring out how to implement them into workflows that meaningfully improves their bottom line, it appears the forward-deployed engineer (FDE) — engineers who work within client organizations to build, implement and deploy software or AI models — is about to become among the most sought-after specialist in the AI industry. 

Demand for FDEs is already rising rapidly, according to the C&T study, which projects demand for these specialists to surge by 2,100% by the end of the year. 

The research draws on interviews with more than 250 C-suite hiring executives across 180 companies, a focused survey of 80 Fortune 500 executives, and interviews with more than 300 FDEs and applied AI engineers between January and June 2026. 

At the start of the year, only 5% to 10% of companies were planning to hire FDEs, and mostly only for small pilots. By the end of the second quarter, however, that number jumped to 70%, with the largest consulting and services firms reporting a need to increase their FDE headcount by 10 times, building full teams of 20 to 100 employees. 

“This is all happening at a speed I’ve never seen. Enterprises are hiring in the middle of summer,” Jeff Christian, founder of C&T told TechCrunch.

That kind of demand will outstrip supply, if C&T’s study is accurate. The report found that there are roughly 17,000 U.S. FDEs on the market today, a good chunk of whom are already employed by Palantir, which invented the concept of the FDE years ago. (Christian said some of his clients are even buying Palantir’s technology just so they can access the firm’s FDEs.) 

Only a fraction of the FDEs out in the wild are apparently elite enough to deliver true ROI, which these days is measured as “multiple tens of millions of dollars of ROI impact,” according to Christian. That could manifest as revenue acceleration on the go-to-market side (lead generation) or “replacing FP&A or replacing 2,300 document processors in India,” Christian says. 

As Chris Taylor, CEO of Ode with Anthropic (a new FDE-focused services firm), put it: “Many FDEs are well equipped to help you roll Claude Code out to your workforce. Very few are capable of building your flagship AI product feature.”

Now that token-maxxing has morphed into value-maxxing, and enterprises taking a harder look at their balance sheets, accounting for AI spending is becoming more important than ever.

“This fall, [Wall Street] is about to say, ‘Hey, we’ve given you two years to figure this out…and you haven’t. There’s no ROI. So we’re going to start punishing those that have spent hundreds of millions, maybe even billions on this, and aren’t generating ROI, and rewarding those that have’,” Christian said.

AI companies are under pressure, too, as they’ve already spent tens of billions to train and deploy their models. For frontier AI firms, reaching profitability will depend on whether they can inject their technology into as many enterprises as possible, though that task is now being threatened by cheaper, increasingly capable open-weight models from China.

That’s why firms like OpenAI and Anthropic have set up their own ventures — Ode with Anthropic and OpenAI’s Deployment Company — and staffed them with FDEs whose sole purpose is to go forth and spread their tech around the enterprise. 

It’s not only top AI firms and large consultancies clamoring for FDEs, however. Enterprises from insurance and fintech to healthcare and gaming, are seeking out these specialists, Christian said.  

Companies are hiring teams of FDEs instead of bringing them in from firms like Ode, or Deployment Co, seeking to keep knowledge of proprietary processes in-house and protect them from the likes of OpenAI and Anthropic.

“Everybody’s concerned that if they give up their proprietary business processes, [the AI firms] can compete with them, which is true in many different areas,” Christian said. “So having this muscle internally is so important.”

Taylor said he’s starting to hear the phrase “internal forward-deployed engineers” more often, but his clients aren’t yet asking Ode to put together internal FDE teams for them.

While many an enterprising young engineer might think they have the industry expertise and AI chops to take advantage of what may turn out to be a talent war, Christian warns that the FDE may not always be in demand. 

“Maybe in two years, everything’s automated, and agents are automating agents as opposed to humans automating agents,” Christian said. “That is something that could occur. Hopefully, it doesn’t, and we continue to need these people within companies.”

In the medium-term, he thinks the need for FDEs will shift from enterprise AI to physical AI as companies try to implement things like humanoid robots into their workflows. But within five or 10 years, he says it’s entirely possible that the role of FDE will “go away.”

That may be true for all knowledge work, if AI leaders and CEOs’ vehement predictions come true. While C&T focuses on recruiting for fast-growing industries and hasn’t seen a pullback yet, Christian says more general search firms have definitely experienced a slowdown in recruitment requests. 

Everything to do with AI is growing and in demand, he says. For now.

“I think that there’s absolutely a time soon where we’re going to see an impact in our business,” Christian said.

You can even use your Roku like a PC, if you have the right phone.

Mirror your screen or cast individual apps from Android to Roku

renata colella/Shutterstock

Roku allows Samsung Galaxy owners to use wireless DeX for a full desktop experience

Samsung

Nimble, a New York City-based tech startup VentureBeat previously covered for its efforts to re-invent web search for enterprises by using multiple AI agents to improve accuracy and depth, is taking another step toward its vision of a world in which agents do most of the web searching instead of us typing and reviewing the results manually.

Nimble today launched Web Search Agents, a new retrieval system designed to help AI agents perform more 21% more accurate web research while using significantly fewer tokens — 51% less compared with leading AI search alternatives on comparable, according to the firm.

While Nimble did not disclose its specific benchmarking methodology or competitors evaluated, the results underscore a growing trend in enterprise AI: optimizing retrieval has become as important as improving the underlying language models themselves.

Nimble’s leadership says the product combines self-learning retrieval strategies, proprietary web indexes, and live web access to deliver domain-specific search capabilities that outperform general-purpose web search services for enterprise workloads.

“Our research team built self-learning retrieval algorithms that learn a customer’s domain,” said Nimble CEO and co-founder Uri Knorovich in an interview with VentureBeat. “They find the exact information more efficiently, reduce the amount of multi-hop reasoning required, and lower token usage while improving accuracy.”

Rather than positioning itself as another general search engine, Nimble is targeting developers building autonomous agents that require continuously updated information from the public web for research, lead generation, competitive intelligence, compliance, and other business-critical workflows.

It’s also designed to slot in seamlessly to an enterprise’s existing systems and workflows.

“You can run the agent directly through the Nimble API with zero infrastructure,” Knorovich said. “For large enterprises, we’re partnering with Microsoft, Oracle, Snowflake, and others so customers can deploy these agent systems inside their own infrastructure.”

How does it work and stack up to other, existing AI-powered search and agentic systems? Read on to find out.

Moving beyond generic AI web search into specialized search agents that fit your enterprise’s needs

Most AI applications today rely on general-purpose search application programming interfaces (APIs) for search engines and public knowledge bases that return broad collections of files, leaving the language model responsible for determining which sources are relevant.

That process often requires multiple retrieval steps, additional reasoning, and significant token expenditure before an agent produces an answer. This is obviously inefficient and raises the cost spent to run AI search looking through irrelevant sources.

Nimble argues that before long, every enterprise will need its own methods for searching, retrieving, and validating external information since each enterprise relies on its own distinct preferred sources, signals, and standards of trust.

As such, instead of applying one search strategy to every workload, Nimble’s Web Search Agents are designed to learn the characteristics of a specific domain and adapt how information is retrieved, providing agents with structured, relevant context rather than forcing them to sift through large amounts of generic search results.

“Instead of one generic retrieval model, we build specialized retrieval models for each customer’s domain, making them faster, cheaper, and more accurate,” Knorovich explained. “A single enterprise can run hundreds of different agents. Each one has its own domain expertise, guardrails, goals, and search algorithm. The optimization starts with the second search, without requiring any setup from the customer.”

Its goal is not only to reduce redundant retrieval, but also to shorten multi-step research paths and avoid repeatedly sending raw pages through a language model for parsing, resulting in the 51% reduced token figure the company cites.

The distinction is particularly relevant for long-running enterprise agents performing research over hours or days rather than answering simple consumer questions. In those scenarios, reducing unnecessary tool calls can significantly lower operating costs while improving answer consistency.

That emphasis reflects a broader shift occurring across the AI tooling ecosystem. As foundation models become increasingly capable, infrastructure vendors are competing on everything surrounding the model—including retrieval, orchestration, memory, observability, and governance.

Optimizing retrieval for production AI

The launch builds on Nimble’s broader strategy of becoming an enterprise web intelligence platform rather than simply a web scraping provider. Earlier this year, the company introduced its broader Agentic Search Platform following a $47 million Series B financing, positioning itself as infrastructure that transforms the live web into structured, machine-readable data for AI systems.

The company’s latest release extends that vision with a concept it calls “Harness as a Tool,” which powers its new domain-specialized Web Search Agents. Rather than requiring engineering teams to assemble separate search APIs, browser automation, extraction pipelines, validation logic, memory systems, and orchestration code, Nimble packages those capabilities behind a managed interface.

The harness can determine what to search, navigate pages when conventional indexes are insufficient, extract relevant information, validate the results, and return the final context in a form designed for downstream agents.

Nimble also says the system retains domain-specific memory and builds proprietary indexes that improve as customers run more searches.

“The biggest research breakthrough is adding semantic memory and a caching layer to the agent,” Knorovich told VentureBeat. “The agent learns usage patterns and domain expertise over time, so every subsequent search becomes faster and more efficient.”

As for what domains Nimble can tackle, the company says it can address virtually any knowledge work domain.

“We’ve seen customers build investment banking analysts, competitive intelligence agents for product managers, go-to-market research agents, newsroom monitoring, insurance applications, life sciences research, and supply chain optimization,” Knorovich said. “Our customers surprise us every day with new agent use cases.”

However, for enterprises concerned about data privacy and retention, Knorovich assured VentureBeat that: “Nimble is zero-data-retention by design. Customer queries are never stored in our environment, and when customers deploy semantic memory and self-learning models, that knowledge stays in their own tenant—not ours.”

Customer deployments point to operational gains

Nimble supported the announcement with early customer examples from AI-native software vendors and enterprise users.

AI-native CRM company Rox reported achieving a 20× reduction in token costs after adopting Nimble’s retrieval infrastructure while simultaneously improving the quality and completeness of information available to its AI agents.

Although the company did not disclose detailed workload measurements or a reproducible baseline, the example illustrates the operational savings retrieval optimization can provide for high-volume agent deployments.

Nimble says its infrastructure currently supports more than 90 million searches each day across Fortune 500 enterprises and AI-native companies operating mission-critical workflows where accuracy, completeness, and enterprise control are essential.

API, SDK and MCP support target AI builders

The platform is immediately available through an API, SDK, and Model Context Protocol (MCP) integration, allowing developers to connect Nimble directly into AI agents regardless of the orchestration framework they use.

Developers can use the platform for several categories of web intelligence, including:

• Low-latency live web search

• Deep multi-step web research

• Web crawling

• Structured dataset generation

• Domain-specific information retrieval

The company also provides documentation and pre-built agents for common web extraction tasks while allowing developers to build custom retrieval agents using natural-language descriptions instead of manually maintaining scraping logic.

Nimble is offering two notably different consumption models. Developers can begin with a pay-as-you-go Agent API priced from $0.025 per Web Search Agent request at the listed low-effort setting. Companies that want Nimble to configure and manage custom data delivery can instead buy annual managed plans beginning at $2,500 per month.

Where Nimble fits in the emerging agentic search stack

Nimble enters a market that has rapidly expanded beyond traditional web search into autonomous research agents capable of planning, browsing, reasoning, and synthesizing information. Products such as ChatGPT Deep Research, Google Gemini Deep Research, Alibaba’s Tongyi DeepResearch, Perplexity, and Sakana Marlin all seek to automate knowledge work that previously required hours—or, in Marlin’s case, potentially weeks—of human research.

Rather than competing head-to-head as another end-user research assistant, however, Nimble is positioning itself one layer lower in the AI stack—as the web intelligence infrastructure that powers those agents or custom enterprise applications built on leading foundation models.

That distinction reflects an increasingly important architectural shift in enterprise AI. Most “Deep Research” systems optimize the overall research workflow, generating search plans, iteratively gathering information, and producing synthesized reports.

Nimble instead argues that the retrieval layer itself has become the primary bottleneck for enterprise AI deployments. If an agent retrieves too many irrelevant pages or performs unnecessary search iterations, token consumption, latency, and operating costs all increase before the model even begins its main reasoning process.

“Customers across life sciences, insurance, healthcare, pharma, retail, and digital-native companies are all telling us the same thing: we need to feed our agents with more accurate context, and we need to reduce the amount of tokens every task consumes,” Knorovich said.

The launch blog makes that argument more concrete by describing how teams frequently rebuild the same retrieval stack themselves. A production agent may start with a search API, then accumulate browser controls, parsers, extraction components, validation steps, memory, caching, evaluations, and custom workflow logic. Nimble is positioning its harness as a managed alternative to that growing engineering burden.

In Nimble’s view, improving retrieval before reasoning begins is more valuable than simply giving a language model more documents to analyze. The company’s Web Search Agents therefore adapt retrieval strategies to a particular workload, combining proprietary indexes with real-time web retrieval and task-specific search policies rather than applying the same search algorithm across every domain.

That makes Nimble less of a direct competitor to OpenAI’s or Google’s research assistants than to developer-focused retrieval infrastructure such as Exa and Tavily. Those platforms also provide AI-native search APIs and research capabilities, but Nimble differentiates itself by emphasizing self-learning retrieval strategies, proprietary indexing, enterprise governance, managed delivery, and token efficiency for production agents.

For organizations building their own AI systems, the distinction could become increasingly important. Foundation models are becoming more capable across the industry, shifting competitive differentiation toward the infrastructure surrounding them—including retrieval, orchestration, memory, observability, and governance. Nimble’s strategy reflects that broader trend, betting that better web intelligence can deliver larger operational gains than incremental improvements in model reasoning alone.

Enterprise infrastructure versus AI research assistants

The different positioning is also reflected in pricing.While consumer-facing AI research assistants are generally sold as productivity subscriptions for individual users or teams, Nimble is pricing its managed service as enterprise infrastructure designed to power production applications. Its pay-as-you-go API, however, gives developers a lower-cost path to test the underlying agent technology before committing to a managed deployment.

The comparison reveals three increasingly distinct markets.

1. ChatGPT Deep Research, Gemini Deep Research, and Perplexity operate primarily as user-facing research assistants.

2. Exa and Tavily provide developer-facing retrieval and research APIs.

3. Nimble and Sakana Marlin occupy more enterprise-oriented territory, but at different layers: Nimble supplies retrieval infrastructure, while Marlin performs long-horizon strategic analysis.

Sakana Marlin is particularly useful as a counterpoint. It is positioned as a “Virtual CSO” rather than a search API, running autonomous research loops for as long as eight hours and producing executive-ready reports, references, and supporting materials.

Nimble, by contrast, is designed to sit beneath those kinds of systems, supplying the specialized retrieval, browsing, extraction, validation, and orchestration that enterprise agents need to gather reliable external information before reasoning begins.

The comparison therefore should not be read as a direct price-to-price evaluation. A $20/month ChatGPT Plus or $19.99/month Google AI Pro subscription buys an individual AI workspace with Deep Research capabilities.

Nimble’s $2,500/month managed plan funds concurrent production agents, managed ETL, MCP integration, web-page capacity, storage, and hands-free data delivery.

Sakana Marlin’s approximately $936/month (¥150,000/month) Pro plan pays for extended, compute-intensive strategic research workflows.

Each price reflects a fundamentally different product boundary and deployment model rather than simply a different level of AI capability.

Why retrieval is becoming the next AI battleground

As enterprise AI systems mature, the industry is increasingly recognizing that model quality alone does not determine application performance.

Large language models frequently fail not because they cannot reason, but because they lack timely, trustworthy external information. That reality has fueled rapid investment across retrieval-augmented generation, AI-native search, web intelligence platforms, knowledge graphs, browser automation, and agent infrastructure.

Nimble’s launch reflects this evolution by focusing less on building another frontier model and more on improving the quality of information flowing into existing ones.

Whether the company’s reported 21-point improvement in answer quality and 51% reduction in token usage hold up across a broad range of enterprise deployments remains to be independently validated.

The larger strategic bet is that, as frontier models become more interchangeable, companies will differentiate themselves through the data, retrieval policies, trusted-source rules, memory systems, and orchestration layers surrounding those models. Nimble is not trying to build the researcher that sits in front of the user. It is trying to become part of the infrastructure that determines what the researcher can find, how efficiently it can find it, and whether the resulting evidence is complete enough to support production decisions.

Web Search Agents are available through Nimble’s API, SDK, and MCP integrations, with a free trial available for developers evaluating the platform.