Techin' Geek

Leonid Radvinsky, the billionaire proprietor of the adult content platform OnlyFans, has died at the age of 43 following a struggle with cancer. 

OnlyFans confirmed Radvinsky’s demise on Monday. The firm expressed that it was “deeply saddened” by his loss, as stated by a representative, noting that his family has requested privacy. Reuters was the first to break the news. 

Hailing from Odesa, Ukraine, Radvinsky relocated to Chicago as a young boy and began operating adult streaming sites during his teenage years, launching MyFreeCams in 2004.

In 2018, Radvinsky acquired a 75% share in Fenix International Limited, the parent company of OnlyFans, and held the positions of director and majority owner. Aside from OnlyFans, he also invested in technology companies through Leo, a venture capital fund he founded in 2009.

OnlyFans was established in 2016 by Tim Stokely and saw immense growth during the COVID-19 pandemic. The platform allowed creators to earn directly from their work, drawing many individuals from the adult industry looking for a dependable income source. To date, OnlyFans has distributed over $25 billion to its creators.

Radvinsky’s death occurs just months after the company was reportedly in talks to sell a 60% stake in OnlyFans, which would have positioned the firm’s valuation around $5.5 billion. 

There has been extensive discussion about establishing context for AI systems. In the realm of consumer software, startups have emerged focusing on search, documents, and meetings. Their aim is to capture context from your digital activities, create links to additional tools, and allow you to query that information. Some tools have advanced further. For example, Rewind (which transitioned to Limitless and was acquired by Meta) and Microsoft Recall seek to log everything occurring on your screen and assist you in recalling it all.

A fresh startup named Littlebird is pursuing a comparable goal but with a slightly different method. Unlike applications like Rewind that preserve screenshots or some form of visual data, Littlebird is “reading” the screen and logging the context in a text format.

The fundamental concept driving the product is that, by continuously reading your screen, you do not need to provide extra context for productivity. The startup asserts that while many AI tools aim to divert your attention, Littlebird operates quietly in the background, surfacing only when you desire it to.

Upon installing Littlebird on your device, you can specify which applications you want it to overlook and avoid capturing context. The startup claims that it automatically disregards password managers and sensitive input fields in internet forms, such as passwords and credit card information. You can also link additional applications like Gmail, Google Calendar, Apple Calendar, and Reminders to the app.

The application allows you to inquire about your data, providing pre-made prompts to initiate your queries, such as “What have I accomplished today?” or “Which emails are significant to me?” Within a few days of utilizing the app, I found that these prompts became increasingly tailored as time progressed.

Littlebird features an integrated notetaker similar to Granola that utilizes system audio and operates in the background to transcribe meetings and generate notes and action items. When viewing a meeting in detail, there’s an option labeled “Prep for meeting” that considers previous meetings, emails, and company history to furnish you with further insights about the upcoming meeting. This feature also gathers information from platforms like Reddit to inform you about public sentiment regarding a specific product or company.

Another feature named Routines provides detailed prompts for Littlebird to execute at specified intervals, including daily, weekly, or monthly. The company offers several pre-configured routines like daily briefings, weekly activity summaries, and summaries of previous day’s work. Users are also able to design their own routines with tailored instructions.

Littlebird was established by Alap Shah, Naman Shah, and Alexander Green in 2024. Siblings Alap and Naman previously launched Sentieo, a platform aimed at institutional investors, which was sold to the market intelligence company AlphaSense. They also co-founded a wellness food venture called Thistle. Alap was a co-author of the widely discussed Citrini paper on how AI agents could lead to economic disruption, resulting in declines in various tech stocks. Green has founded several enterprises in the fields of hardware, software, and AI.

“We initiated this project when Alap brought up a compelling issue that AI will revolve around the data of [users]. Models lack knowledge about users, limiting their effectiveness. We contemplated different UI and OS paradigms ripe for disruption by AI, and that sparked Littlebird as a project,” Green shared with TechCrunch during a phone call.

Green remarked that while Rewind was somewhat aligned with Littlebird’s goals, it relied heavily on screenshots and did not provide an optimal search experience. He noted that the startup is merely at the beginning of its journey and has a multitude of issues to address, including enabling large language models (LLMs) to grasp various types of user context.

With Littlebird, users have the ability to delete their data at any moment, and their information is stored in the cloud with encryption. Green mentioned that the decision to store data in the cloud was made to facilitate powerful models for various AI processes, which cannot be achieved locally.

“We do not retain any visual data. We solely keep text, making the data significantly lighter. This was likely a factor that contributed to the struggles of Recall and Rewind, as capturing a screenshot requires substantially more data. Furthermore, I believe it is also more intrusive,” he explained.

Littlebird is available for free download and use, but to access enhanced usage limits and features like image generation, users have the option of paid plans starting at $20 monthly.

The startup has secured $11 million in funding, led by Lotus Studio, with involvement from Lenny Rachitsky, Scott Belsky, Gokul Rajaram, Justin Rosenstein, Shawn Wang, and Russ Heddleston.

Many of these investors are active users of the product. Rajaram, who has experience at Google and Facebook within advertising, stated that the product alleviates the difficulty of recalling, accessing, and re-explaining one’s own work. DocSend co-founder and CEO Heddleston shared that he revised the firm’s marketing website using the tool, leveraging context from meetings, emails, Notion, and more.

Rachitsky, who manages his own newsletter and podcast, remarked that AI’s effectiveness is linked to the context it possesses, acknowledging that it frequently overlooks significant elements of your day. He mentioned that he utilizes the tool to enhance his productivity strategies and to foster greater happiness. For continued success, he suggested that the product would need to discover a major use case.

“I believe it is crucial to identify that essential use case. That is what determines this product’s current success. Numerous individuals have already identified that for themselves, and the team is attuned to these experiences as they recognize these use cases developing,” he observed.

“I’ve hosted numerous AI product creators on the podcast, and a recurring theme is that you won’t truly understand how users will interact with your product until it is released. The strategy is to launch initial versions, observe user interactions, and focus on those use cases instead of waiting for everything to be perfectly defined.”

Zain Asgar, an adjunct professor at Stanford and a successful entrepreneur, has secured an $80 million Series A funding for a startup addressing the AI inference bottleneck issue in an insightful fashion. Menlo Ventures led this investment round. 

The startup, Gimlet Labs, claims to have developed the first and only “multi-silicon inference cloud,” which is software enabling simultaneous execution of AI workloads across various hardware types. It can distribute an AI application’s tasks among both conventional CPUs and AI-optimized GPUs, as well as high-memory architectures.  

“In essence, we operate across all available hardware types,” Asgar shared with TechCrunch. 

One agent may connect several steps together, each requiring distinct hardware: Inference is compute-bound; decoding is memory-bound; and tool calls are network-bound, explains lead investor, Tim Tully of Menlo, in a blog post regarding the funding.  

No single chip currently does it all, but as new hardware is introduced and older GPUs are repurposed, “the multi-silicon fleet is prepared — it merely needs the software layer to function.” This is what Tully believes Gimlet Labs provides.

If the ongoing trend of deploying more computing resources persists, McKinsey predicts that spending on data centers will reach nearly $7 trillion by 2030. Asgar mentions that existing applications are utilizing the current hardware deployed “only between 15 to 30 percent” of the time.  

“Another perspective is that you’re wasting hundreds of billions of dollars by permitting resources to sit idle,” he commented. “Our goal was essentially to determine how to make AI workloads 10x more efficient than ever before, today.” 

As a result, he and his co-founders, Michelle Nguyen, Omid Azizi, and Natalie Serrino, began to develop orchestration software that breaks down agentic workloads, allowing them to be concurrently distributed across various hardware infrastructures. 

Gimlet Labs asserts that it can enhance AI inference speed by 3x to 10x without increasing cost or power consumption. Gimlet claims it can even partition the underlying model to run across different architectures, selecting the optimal chip for each segment of the model. 

The firm has established partnerships with chip manufacturers NVIDIA, AMD, Intel, ARM, Cerebras, and d-Matrix.  

Gimlet’s offering, available as software or via an API to its Gimlet Cloud, is not intended for the general AI application developer. It targets the largest AI model laboratories and data centers. 

The company officially launched in October, reporting eight-figure revenues right from the start (at least $10 million). Asgar noted that their customer base has more than doubled in the past four months and now includes a significant model manufacturer and an extremely large cloud computing firm, though he opted not to disclose their names.  

The co-founders had previously collaborated at Pixie, a startup that developed an open-source observability tool for Kubernetes. Pixie was acquired by New Relic in 2020, just two months post-launch with a $9 million Series A led by Benchmark. (Pixie’s technology is now part of the open-source organization that manages Kubernetes.)  

After Asgar coincidentally met Tully about a year ago and also secured angel investments from Stanford faculty, venture capitalists began reaching out. Following the launch, a term sheet arrived on Asgar’s desk. When VCs discovered that Asgar was evaluating offers, “we received a significant influx of funding,” and the round was quickly oversubscribed, he stated. 

With the prior seed funding, the startup has now raised a total of $92 million, backed by numerous angels, including Sequoia’s Bill Coughran, Stanford Professor Nick McKeown, former VMware CEO Raghu Raghuram, and Intel CEO Lip-Bu Tan. The company currently has a workforce of 30 people.

Other investors consist of Factory, which led the seed funding, Eclipse Ventures, Prosperity7, and Triatomic.

The delivery powerhouse Grab announced on Monday its intention to acquire Delivery Hero’s Foodpanda operations in Taiwan for $600 million in cash, representing its initial growth beyond Southeast Asia. Grab mentioned that the transaction is subject to regulatory approval and is expected to finalize in the latter half of 2026. The firm plans to complete the transition of users, merchants, and driver-partners to its platform by early 2027.

This decision follows approximately a year after Uber Technologies withdrew its plan to acquire Foodpanda’s operations in Taiwan in March 2025, after the transaction was obstructed by Taiwan’s antitrust authority due to competition issues.

Previously, Uber Eats and Foodpanda were leading players in Taiwan’s food delivery sector. A recent study indicated that Foodpanda commanded a 52% market share, while Uber Eats represented 48% between 2022 and 2023. Taiwan’s Fair Trade Commission stated that the merged entity would have dominated approximately 90% of the market, raising alarms about decreased competition and possible price hikes.

However, this scenario may demonstrate a different situation. Should Grab secure Foodpanda’s Taiwan business, the Singapore-based ride-hailing and delivery company would achieve a market share of just over 50%, positioning itself as a more robust rival to Uber Eats instead of establishing a near-monopoly.

“This represents a natural progression for Grab, as our experience in Southeast Asia aligns perfectly with this market. Our extensive know-how in managing intricate delivery logistics for densely populated and high-traffic urban areas is ideally suited for Taiwan’s vibrant cities,” Anthony Tan, Group CEO and co-founder of Grab, stated in the announcement. “Taiwan’s populace of around 23 million also reflects a strong demand for mobile-first services, akin to the Southeast Asian consumers that Grab services daily. We perceive a considerable opportunity to expand the food and grocery delivery landscape here.”

Following the acquisition, Grab plans to extend its reach to 21 cities throughout Taiwan, fortifying its position in a crucial market. The agreement integrates Grab’s AI-enhanced platform and operational expertise with Foodpanda’s extensive local presence. Foodpanda’s operations in Taiwan produced about $1.8 billion in Gross Merchandise Value (GMV), as reported by the company.