Former Plaud executive raised $26 million in financing for the backend monitoring AI application
TechCrunch
1h ago
Ai Focus
The Peter Mo-founded Ambi is developing an application that continuously listens to the user environment. The company claims to have raised $26 million in funding and plans to launch iPhone, Mac, Apple Watch applications, as well as its own hardware, but their approach has also raised concerns regarding privacy.
Helpful
No.Help

One of the core commitments of AI is that the more it knows about a person, the better it can assist in completing tasks and provide appropriate advice. And if you believe what tech companies say, this requires applications and devices to constantly listen to, observe, or record user activities in order to collect as much contextual information as possible.

Obviously, continuous collection of user data raises concerns regarding privacy and data security. However, startups are still advancing such applications without any impact, firmly believing that the personalized assistant market is large enough to accommodate these concerns.

The former Plaud executive and the current Peter Mo are working on creating Ambi, doing something similar. The goal of this application is to provide users with real-time intelligence and insights by constantly listening to them.

The company stated on Monday that it had raised $26 million in multiple rounds of financing, led by Luminous Ventures. Luminous Ventures has invested in companies such as Meituan and Yushu Technology, and ODM manufacturer Huacin also participated in the investment.

Ambi will also launch iPhone, Mac, and Apple Watch applications. These applications function similarly to meeting note-taking tools, but users can choose to have them listen continuously in order to capture the context throughout the day.

The meeting recording company has been trying to automate a series of tasks that follow a meeting. These tasks are based on the summaries and action items collected during the meeting, and are carried out by connecting various software systems through numerous connectors. Ambi hopes to take it a step further by enabling their agents to start working even during the meeting itself.

The company claims that its AI intelligent entity can proactively offer suggestions or retrieve relevant knowledge during meetings based on the user's calendar, meeting schedules, knowledge base, and web page data. Users can also start tasks on one device and continue them on another.

Mo to TechCrunch says: "The way we interact with intelligence is limited. Many times, intelligence is confined to the screen; if you use hardware like Plaud, you have to wait for the recording to be sent to your phone, uploaded to the cloud, and then processed before you can get intelligent results. This is why we created Ambi to make AI available anytime, anywhere."

The company has even greater ambitions for information collection. Its upcoming desktop application will be able to monitor what users are working on and store important parts of that information in its memory, in order to provide more personalized suggestions. Applications like Littlebird or Minimi are already doing similar things.

This startup also plans to launch its own hardware devices, equipped with screens, speakers, cameras, and microphones, to capture context and display information.

Privacy Issues

The release of Ambi coincides with an increasing concern from the public regarding AI devices, which often lack sufficient indications to let users know that they are collecting data on the people around them. It is still unclear whether Ambi applications or upcoming devices will provide adequate warnings to inform third parties that their conversations are being recorded, in cases where users do not proactively disclose that they are using Ambi technology.

Some meeting recording tools have even been sued for recording others' conversations and data without consent. In addition, Apple's recently launched Voice Intelligence feature, which continuously transcribes voice, has also raised privacy concerns.

Mo states that users can disable the environmental data collection function in the Ambi app and device at any time, and that this function is disabled by default. The company claims it does not store user recordings and encrypts the data, but did not specify the encryption method used. Despite these measures, the responsibility for disclosing environmental data collection still rests with the user. He considers this a problem faced by the entire industry.

He said, "We must tell users that they need to explain to others about the environmental data collection. As an industry, we haven't yet solved this issue [of disclosure]. We have included as much information about disclosure as possible in various places, but if users try to hide their collection activities, we can't just resolve the problem that way."

Ambi currently has offices in San Francisco, Beijing, Shanghai, and Shenzhen, with over 50 employees. The company is recruiting for positions responsible for market expansion as well as AI engineers who work on speech recognition models.

Tip
$0
Like
0
Save
0
Views 24
HQYC reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
Microsoft plans to deploy its own chips on a large scale in data centers, with a capacity of several gigawatts
Microsoft stated that it plans to deploy Maia chips with a total capacity of several gigawatts in its data centers, among which the latest generation of Maia 300 will begin deployment next year. These chips are specifically designed for AI inference, and Microsoft claims that this will enable more efficient operation of AI inference and reduce some of the dependence on NVIDIA GPU.
The Block
·2026-10-12 22:12:31
14
XRP Trading volume plummeted by 79% on the occasion of a price rebound; over $40 million in long positions were forcibly liquidated
From October 8th to 10th, the daily trading volume on Binance’s XRP / USDT decreased from $252.45 million to $54.12 million, a decline of about 78.6%. During the same period, the closing price of XRP rose by 1.6%. Prior to this, there was a wave of liquidations in the XRP derivatives market. Data from XRPL.to shows that from October 7th to 8th, approximately $40.7 million in long positions on Binance, Bybit, and OKX were forcibly closed, while short liquidations amounted to only $1.72 million.
Coinpaper
·2026-10-12 22:04:14
19
Is Bitcoin " Uptober " at Risk in 2026? ETF Capital Outflow and Fed Rate Resetting Will Create Pressure Before October 28
In early October 2026, Bitcoin continued its historical statistical advantage of " Uptober ", but after the release of the Federal Reserve's September meeting minutes, the price fell below $83,000, and spot Bitcoin saw a net outflow of $484.9 million in a single day. The article states that the FOMC meeting on October 27-28, the PCE inflation data on October 29, and the subsequently released CPI will determine whether the seasonal trend of October can continue.
Coinpedia
·2026-10-12 22:04:12
20
RLUSD of Ripple has received support from Standard Chartered Bank, and a $1.5 billion cooperation has been launched to advance LMAX.
Standard Chartered Bank has launched an institutional RLUSD custody service on XRP Ledger, advancing the cooperation between Ripple and LMAX Group. The three companies are working on a scheme that allows financial institutions to use RLUSD as transaction collateral without the need to transfer tokens directly into the trading venues.
Coinpaper
·2026-10-12 22:04:10
17
Report: Microsoft to Deploy Self-Developed AI Chips on a Large Scale Next Year, Aiming to Improve Inference Efficiency by 40%
According to Bloomberg, Microsoft announced that a new generation of Maia 300 chips will begin to be deployed in data centers next year. The company plans to use these self-developed chips on a gigawatt scale and expects that the number of token that can be processed per megawatt will be 40% higher than that of mainstream GPU systems. Microsoft currently has no intention of renting out the computing power of the Maia chips to external customers, preferring to meet its own AI service needs first.
Wallstreetcn
·2026-10-12 22:04:09
21
View More