Topics
Latest
AI
Amazon
Image Credits:Nuthawut Somsuk / Getty Images
Apps
Biotech & Health
Climate
Image Credits:Nuthawut Somsuk / Getty Images
Cloud Computing
Commerce
Crypto
Read’s platform taps generative AI to summarize meetings, messages and emails.Image Credits:ReadImage Credits:Read
Enterprise
EVs
Fintech
Image Credits:ReadImage Credits:Read
fund raise
Gadgets
stake
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
societal
Space
Startups
TikTok
exile
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
video
Partner Content
TechCrunch Brand Studio
Crunchboard
get hold of Us
Meetings are prison term - consuming , and there ’s no mode around it . According to a 2022 poll from Deputy.com , manyU.S. workers spend up to around eight hours in meetings every calendar week , count on the manufacture and locale .
The productiveness hit explain the growing popularity of AI - powered summarization creature . In a recent survey of vendor by The Conference Board , a nonprofit think tank , nearly half of answerer said they were using AI to summarizethe content of email , conference calls and more .
While a number of videoconferencing suites now propose work up - in summarization features , David Shim believes that there ’s room for third - party solutions . And he would : He ’s the co - founding father ofRead AI , which summarizes television calls across political program such as Zoom , Microsoft Teams and Google Meet .
Shim , antecedently the CEO of Foursquare , co - founded Read AI with Rob Williams and Elliott Waldron in 2021 . Prior to Read AI , the threesome worked together at Foursquare , Snapchat and Shim ’s previous inauguration , Placed ( which Foursquare acquired in 2019 ) .
“ Read AI ’s lineal rivalry is traditional project management , where notes are manually written , ” Shim separate TechCrunch . “ By learning what ’s important to you crossing - platform , Read is n’t a carbon monoxide gas - pilot — rather , it ’s an automatic pilot delivering content that crap your work more good and effective . ”
At the head start , Read concenter solely on video meetings solvent , offeringdashboardsto measure how well a get together ’s become ( as judge by sealed metrics , at least ) and two - minute sum-up of hourlong meeting . But , co-occur with a recently closed $ 21 million funding circle conduct by Goodwater Capital with Madrona Venture Group , the company is expanding into message and email summarisation .
usable in “ easy launch , ” Read ’s raw capability connects to Gmail , Outlook and Slack as well as videoconferencing platforms to learn issue that might be relevant to you . Within 24 hours of connecting to the electronic messaging and videoconferencing Robert William Service you use , Read begins delivering daily updates with summaries , AI - generated “ takeout , ” an overview of key content and update to conversation topics in chronological club . Read charge a $ 15 to $ 30 monthly fee for its service .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
“ What realize Read unique is that its AI agents play quietly in the background , enabling your get together , electronic mail and messages to interact with each other , ” Shim said , adding that the average summary from Read AI condenses 50 electronic mail across 10 recipient into a single summary . “ This connected intelligence unifies your communications and empowers you and your team with personalize , actionable briefings tailored to your needs and priorities . ”
Now , color me unbelieving , but I ’m not certain I trustanyAI - driven tool to summarize content consistently accurately .
Models like ChatGPT and Microsoft ’s Copilotmake mistakes when summarizingbecause of their tendency tohallucinate , including insummaries of meetings . In a late piece , The Wall Street Journalcitedan instance where , for one early adopter using Copilot for meeting , Copilot invented attendees and implied that calls were about study that were never actually discussed .
Is Read AI ’s tool any different ? Shim claims that it ’s more robust than many of the solutions out there , admit rivals likeSupernormalandOtter .
“ Read runs a proprietary methodology to organize naked contentedness with terminology model end product , so that deviations are automatically detected and fittingly steered , ” he suppose . “ to boot , we can practice content from meeting to better contextualize e-mail and messaging content , further reducing uncertainty and meliorate results . ”
Take that statement with a metric grain of saltiness . Shim did n’t share benchmark results to support those assertions .
In lieu of benchmarks , Shim emphasized the productiveness boost summarization tool such as Read can ( in theory ) deliver .
“ Rather than reschedule a meeting as you ’re running belated or twice - booked , Read can go to in your place and deliver to you a sum-up and action items that even the best executive supporter could n’t match , ” he enjoin , stressing also that Read does n’t utilize customer data to train its AI models and that users have “ full control condition ” over contentedness passing through the platform . “ AI is bring focusing back to knowledge workers [ by ] save them minute a mean solar day . ”
Read AI is no alien to controversy , so it ’s alittlehard to take Shim at his Word of God . The platform ’s persuasion analysis tool , which interprets meeting participants ’ vocal and facial cues to inform host on their sentiment , has beencalledout by privacy advocates for beingoverly invasive , prone to bias and very peradventure a data protection hazard .
Gender and racialbiasesare awell – documentedphenomenoninsentimentanalysisalgorithms .
Emotional depth psychology model lean to assign morenegative emotionsto Black people ’s face than white masses ’s , andperceivethe language that some Black people utilise as fast-growing or toxic . AI video hiring platforms have beenfoundto respond otherwise to the same occupation candidate tire unlike rig , such as glasses and headscarves . And in a 2020studyfrom MIT , researchers showed that algorithms could become biased toward sure facial grammatical construction , like smile , which could cut down their accuracy .
Perhaps tellingly , Shim continues to see Read ’s persuasion analysis technology as acompetitiveadvantage , not a risk , while pointing out that client can invalid the feature and that analysis data is erase from Read ’s server periodically . “ Using a multimodal model allows Read to incorporate non - verbal reaction into group meeting sum-up , ” he said . “ As an deterrent example , during a pitch meeting , a startup might speak about the benefits of the Cartesian product , but the participants visually shake their head and frown during the pitch shot … Read create a custom service line of engagement and sentiment for each meeting participant , rather than applying a one - sizing outfit all model , ascertain that each person is treated as a alone person . ”
precise or no , with a $ 32 million war chest and a client base that grew by half a million users over the past quarter , Read clear has some folks win over that it can deliver on its promise .
Read , found in Seattle , Washington , plans to double over its staff to over 40 employees by the final stage of the year leveraging the unexampled infusion of capital , Shim said .
“ In font of a broader retardation over the last few years , Read has continued to see the growth curvature steepen across users , encounter and revenue , ” he added . “ This speedup in increment can directly be attribute to the quantifiable return users see in term of time savings when using Read AI in their meeting . ”
