1663323011-logo2022.png
SIGN IN

Google DevFest 2026 | Talk Data, Meet Developers, and Pick Up Some Swag!

1750128659-英文logo带背景

Posted at 7 hours ago

Developers, are you ready?

The annual Google DevFest is just around the corner!

This year, MagicHub.com will join Magic Data at Google DevFest 2026. We’re looking forward to meeting developers, tech enthusiasts, and open-source collaborators from across the community.

We’re bringing plenty to talk about: speech technology, open-source resources, and a few gifts to take home.

01 | Let’s Talk Tech: Explore Our Mandarin and Chinese Dialect Duplex Speech Data

What changes when AI no longer has to wait for you to finish speaking?

Traditional voice interactions usually follow a “you speak, then I speak” pattern. In everyday conversation, though, people interrupt, jump in, give brief responses, and speak at the same time.

For voice agents to support more natural conversations, models need to recognize speech accurately and generate natural speech. They also need to understand and respond to what’s happening in real time.

High-quality duplex speech data is essential to making this possible.

At DevFest 2026, we’re showcasing 100,000+ hours of Mandarin and 10,000+ hours of multi-dialect duplex corpora. Stop by our booth to explore the datasets, listen to samples, and chat with our team!

We’d also love to discuss the data challenges you’re facing in development:

  • Voice agent: What data do real-time voice agent need? How can they handle interruptions, competing turns, and overlapping speech?
  • Model training and evaluation: How can duplex data support the development and improvement of automatic speech recognition (ASR), text-to-speech (TTS), spoken language understanding (SLU), and voice LLMs?
  • Data engineering: How can multilingual duplex data, custom data collection and annotation, and appropriate licensing and permissions fit into your existing model training workflows?

Whether you’re building a voice product or exploring full-duplex technology, bring your questions and join the conversation.

02 | Let’s Talk Careers: We’re Hiring Data Engineers and AI Trainers

Better models depend on the engineering work behind the data. From collecting and cleaning speech data to designing annotation frameworks and evaluating quality, every step affects a model’s performance.

We’re looking for people who are passionate about AI data to join us in roles including data engineers and AI trainers.

If you’re curious about how data shapes model capabilities and want to work on real-world AI data projects, stop by our booth for a chat.

03 | Let’s Talk Open Source: A Shared Interest Grows into a Community

Struggling to find the right data—or to make it work for your project?

To make speech data easier for AI developers to access and use, Magic Data launched the MagicHub.com in April 2021. The goal is to help more people discover and use high-quality speech data for free.

To date, the MagicHub.com community has released over 200 open-source datasets spanning 60 languages and dialects, including Mandarin, regional Chinese dialects, and other languages. These datasets support tasks such as speech recognition, speaker recognition, speech synthesis, and large language model fine-tuning.

Every developer has a part to play in helping the community grow. Whether you already use MagicHub.com or are discovering it for the first time, we’d love to hear about the datasets you need, features you’d like to see, and challenges you’ve encountered.

P.S. This year, the MagicHub Foundation launched the MagicHub Ambassador Program, seeking like-minded developers and student researchers. Drop by our booth to learn more and join the movement!

04 | Open Mic: One Dataset, Different Perspectives

You’ll also find us at this year’s DevFest Open Mic!

An AI voice industry consultant from the MagicHub.com community will give a talk titled:

“One Dataset, Different Perspectives: What Business Leaders, Algorithm Engineers, and Investors See”

How does the value of a dataset change depending on who’s looking at it? Join us to explore the different perspectives people bring to AI data.

⏰ Time: November 7, 2026, 14:30–16:30
📍 Location: Open area, 3rd floor, Jeurong Hotel Shanghai

05 | Pick Up Some Swag: Good Conversations, Fun Gifts

We’ve prepared notebooks, stickers, surprise tea capsules, canvas tote bags, and more. Join our booth activities and take home a gift!

🎁 Activity 1: Talk Voice Agents and Choose a Gift

Scan the QR code to join our event chat group and share your thoughts on voice agents and what you’d like them to do.

Then choose one of the following gifts : A notebook、A set of stickers、A surprise tea capsule

🎁 Activity 2: Connect with Our Community and Get a Tote Bag

Join our event chat group and follow the “麦吉哈” (MagicHub.com) official WeChat account to receive a canvas tote bag!

🎨 Activity 3: DIY Sticker Challenge

Get creative with MagicHub.com stickers!

Mix and match the stickers to make your own design, take a photo, and share it in the event chat group.

After the event, we’ll select our favorite creative entries and award prizes, including custom MagicHub.com mugs!

📍 Where to Find Us

Event: Build with Gemini | Google DevFest 2026
Date: November 7, 2026
Venue: Jeurong Hotel Shanghai, No. 80 Minhong Road, Minhang District, Shanghai
Booth: Magic Data booth

From speech data and voice agent to open-source collaboration and career opportunities, we’re looking forward to sharing ideas and getting to know you.

Working on voice agent or full-duplex speech interaction? Looking for high-quality speech data to train your models? Interested in open-source collaboration or a career in AI data? Or simply hoping to meet people who share your interests?

Come meet MagicHub.com × Magic Data at DevFest 2026. We look forward to seeing you there!

💬 Scan the QR code to join our event chat group and connect with fellow developers before the event!

Related Datasets

Datasets Download Rank

ASR-RAMC-BigCCSC: A Chinese Conversational Speech Corpus
Multi-Modal Driver Behaviors Dataset for DMS
ASR-SCCantDuSC: A Scripted Chinese Cantonese (Canton) Daily-use Speech Corpus
ASR-SCSichDiaDuSC: A Scripted Chinese Sichuan Dialect Daily-use Speech Corpus
ASR-CCantCSC: A Chinese Cantonese (Canton) Conversational Speech Corpus
ASR-SCCantCabSC: A Scripted Chinese Cantonese (Canton) Cabin Speech Corpus
ASR-CShhiDiaCSC: A Chinese Shanghai Dialect Conversational Speech Corpus
ASR-EgArbCSC: An Egyptian Arabic Conversational Speech Corpus
ASR-SCShhiDiaDuSC: A Scripted Chinese Shanghai Dialect Daily-use Speech Corpus