As an example that might be helpful to the rest of the group, I was asked by a colleague to help brainstorm methods of evaluating a chatbot we are developing to help practice communication skills for trainees. I ran a Deep Research search and I was very impressed not only with the depth of research but also the recommended approach to evaluation and summary table. Check out the full conversation here: https://www.genspark.ai/agents?id=295c3832-7dc3-41f6-bc5f-81b5f50c8102

Have you used Deep Research? What did you look for and how did it go?

< Review on the Genspark Hackathon >
Hello, I’m Smith (smithx.eth), a Genspark Genius.
On August 16, we held the Genspark Hackathon at Urban Classic Jongno Branch 1. Although the event lasted only four hours, from 2:00 PM to 6:00 PM, it was an incredibly productive and inspiring experience thanks to the ideas and enthusiasm of all our participants.
I did not prepare this hackathon alone. Genspark reviewers Seo Woo-sung, Dohee, and Anji worked with me throughout the preparation process. Their support in shaping the event and preparing the venue helped us bring everything together successfully.
The theme of the hackathon was:
“Connect Ideas, Turn Possibilities into Reality.”
Even the best idea remains only a possibility until someone takes action. This hackathon was designed to give participants an opportunity to connect their ideas with AI and transform them into tangible outcomes within a limited amount of time.
Four hours is not enough time to build something perfect. That made it even more important to clearly define the problem, identify the core value, and focus on the most essential features. Watching participants turn their ideas into working results in such a short period was inspiring for me as well.
Another distinctive feature of the hackathon was that the judging process was conducted entirely by AI, without human intervention.
Neither I nor the event organizers and reviewers decided the final rankings. AI evaluated the completed projects and selected the first-, second-, and third-place winners. Using AI not only to support project development but also to evaluate the final results made this a particularly meaningful experiment.
At the same time, an AI judging system depends heavily on how its evaluation criteria are designed. Based on what we learned from this event, I hope to further refine both the judging criteria and the overall process for future hackathons.
The participants selected by the AI as the first-, second-, and third-place winners received Genspark carry-on luggage merchandise valued at approximately KRW 70,000. Congratulations to all three winners.
I would also like to applaud every participant who turned an idea into a tangible result within such a limited amount of time, regardless of the final ranking.
There were naturally some concerns while preparing the event. However, seeing participants fully engaged in their projects and exchanging ideas made all the preparation worthwhile.
I believe the true value of a hackathon is not simply in selecting winners. It is about taking an idea that previously existed only in your mind, putting it into action, identifying what can be improved, and finding the motivation to move forward.
Once again, I would like to thank Seo Woo-sung, Dohee, and Anji for preparing the event with me, as well as everyone who joined us and contributed their ideas and energy.
I would also like to express my sincere gratitude to Genspark Marketing Reader @Denny Seo & Genspark Ambassador @최재용 디지털융합교육원 for taking the time to attend and honor us with his presence.
I hope this will not be a one-time event. We will continue working to create more opportunities for people to build, test, and share their ideas.
Connect ideas and turn possibilities into reality.
IDEA · CONNECT · IMPACT
Thank you. 

Hey Genspark Community! 👋

I want to share a project I've been working on over the past few weeks – a fully automated audio mastering pipeline that combines the Genspark API with Groq AI (openai/gpt-oss-120b)! 🎧🤖

🎯 What Does the Workflow Do?

This n8n workflow analyzes audio files completely automatically and calculates professional mastering parameters:

✅ DSP Analysis (Python): Crest Factor, LUFS, RMS, Peak, Spectral Energy
✅ Genspark Research: Searches for professional mastering targets for your genre
✅ 4x Groq AI Nodes (openai/gpt-oss-120b): Calculates EQ, compression, limiting
✅ Multi-Stage Validation: AI checks all values for audio engineering correctness
✅ Gmail Report: Automatically sends you a detailed mastering report

🔥 Why Groq Instead of Genspark LLM?

I used Groq with openai/gpt-oss-120b because:

• 🚀 120B-Parameter Model – significantly more powerful & precise for audio engineering than Llama models

• 📊 Native JSON Mode – perfect for structured outputs (Hz, dB, ratio, attack/release, etc.)

• 💰 100% Free – 14,400 requests/day in free tier (no credit card required!)

• ⚡ ~500 Tokens/s – faster than most LLM nodes

Unfortunately, I couldn't find a comparable high-performance model via the Genspark API for this type of workflow automation – hence the Groq integration! 💡

🎛️ 20+ Audio Parameters & 58 Genres

The workflow calculates professional mastering values:

Dynamics: LUFS, RMS, Peak, Crest Factor, Dynamic Range, Transients
Spectrum: Sub-Bass, Bass, Mids, Highs (Hz distribution)
Processing: 5-band EQ (Hz, dB, Q), Compression (Ratio, Threshold, Attack, Release), Limiting
Metadata: Sample Rate, Bit Depth, Format, BPM, Key

Genre Classification from 58 Predefined Genres:
Hardcore, Frenchcore, Industrial, Hard Techno, Techno, House, Trance, Drum & Bass, Dubstep, Hip-Hop, Rock, Metal, Pop, Cinematic, Ambient, Acoustic, Orchestral, Experimental, Speech, and more.

Genre Customization: You can add or edit genres in the Audio Classification node by modifying the JSON Expression. When the classification detects "Unknown", the AI agents autonomously select the best mastering approach based on the complete DSP analysis. 🎨

🤖 AI Agent Autonomy

The 4 AI agents work completely autonomously:

Each node receives the full DSP analysis and determines independently what needs to be done:

• ✅ Already mastered track? The agents recognize it and avoid over-processing

• ✅ Missing dynamics? The agents calculate optimal compression ratios

• ✅ Frequency imbalances? The agents design corrective EQ curves

• ✅ Excessive limiting? The agents recommend gentler settings

No manual intervention needed – the agents analyze, validate, and adapt based on the audio's actual state! 🧠

🔬 Groq Nodes: Under the Hood

The 4 Groq nodes contain highly specialized audio engineering logic:

⚠️ In the video tutorial, I intentionally skipped the Groq node settings because:

• 300+ lines of system prompts per node (too complex for standard users)

• Proprietary DSP terminology (Crest Factor, PLR, LRA, multiband processing)

• Sensitive data (production targets for Spotify, YouTube, mastering)

✅ BUT: The complete n8n workflow is open-source and fully editable! Audio engineers can view all prompts and customize them for their own mastering needs. 🛠️

📦 Download & Setup

🎥 Video Tutorial: https://www.genspark.ai/aidrive/s/7077d887-d109-4c5d-9cfc-5f5ace19b749/85694443-818f-4bab-81d7-700c29fc8055
📄 n8n Workflow JSON: https://www.genspark.ai/aidrive/s/7077d887-d109-4c5d-9cfc-5f5ace19b749/12ba3c42-2d9a-43f1-aeb9-c550ef604b80
📖 Detailed Setup Guide: https://www.genspark.ai/aidrive/s/7077d887-d109-4c5d-9cfc-5f5ace19b749/85694443-818f-4bab-81d7-700c29fc8055

What you need (all free):

• Genspark API Token (create here)

• Groq API Key (sign up here – no credit card!)

• Gmail App Password (create here)

Setup Time: ~10 minutes
Cost per Run: ~$0.01 (only Genspark API, rest is free)

💡 Use Cases

This workflow is perfect for:

• 🎵 Music Producers – Mastering recommendations for your tracks

• 🎙️ Podcast Editors – Loudness normalization (LUFS targets)

• 🎬 Video Creators – Audio post-production checks

• 🔊 Audio Engineers – Automated quality control

🙏 Feedback & Questions?

If you test the workflow or have questions – feel free to write in the thread! 💬

Looking forward to your feedback! 🚀

Developed by: Demian | Genspark Genius Ambassador for Creative Work
Tech Stack: Genspark API, Groq AI (openai/gpt-oss-120b), n8n, Python, Gmail
Release-Date: July 2026

Jaeyong Choi – Genspark Genius Ambassador Profile

1. Jaeyong Choi is a Genspark Genius Ambassador and an expert in Generative AI and AI Agent education.

2. He currently serves as President of the Korea Genspark AI Research Association, promoting the adoption and practical application of Genspark and AI Agent technologies in education and professional settings.

3. As Director of the Digital Convergence Education Institute, he provides training on the practical use of Generative AI and AI Agents for corporations, public institutions, and educational organizations.

4. He also serves as an Adjunct Professor at the Graduate School of Knowledge Service & Consulting, Hansung University, where he continues his teaching and research in the field of artificial intelligence.

5. He completed the “Generative and Agentic AI” programme at Oxford Saïd Business School, further strengthening his expertise in Generative AI and Agentic AI.

6. He holds an IBM Certified AI Developer credential and focuses on applying AI technologies to education, workplace innovation, and practical business transformation.

7. Through publications including 300× Ways to Leverage Genspark AI in the Age of AI Agents, he has introduced practical approaches to using Genspark and AI Agents.

8. In 2026, he also co-authored Genspark Is All You Need, contributing to making Genspark more accessible and practical for non-developers.

9. Since being selected as a Genspark Genius Ambassador in 2026, he has continued to promote Genspark through Korea Genspark AI Research Association Meetups, professional training programmes, and educational initiatives.

10. Jaeyong Choi is committed to building a Genspark-based AI education ecosystem guided by the principle: “Go beyond knowing AI—turn AI Agents into real-world results.”

    

He enjoys sharing practical AI use cases, creating educational content, and helping people explore how AI can improve their work and creativity. Through community events, workshops, and collaborations, he's working to grow the Genspark community across India.

If you're in India or interested in AI, Genspark, or collaborating on events, feel free to connect with Jigar.

LinkedIn: https://www.linkedin.com/in/jigarvyass/

Before Genspark, deal information would come in through a form, land in Slack, and sit there waiting for someone to manually process it. Now Goose handles the whole flow: when an agent closes a deal, Goose posts a Slack message with claim buttons, updates the commission sheet, logs the override, and posts to the team leaderboard. The core setup took about 30 to 40 minutes.

Seve pays roughly what he'd pay a contractor, but says Goose has greater context than any human could, and it can code and set up its own webhooks. His next idea is a second agent called Rooster that handles the team and escalates to Goose when needed. "We'd be ten times more productive," he said.

Read the full story → https://www.genspark.ai/blog/entrepreneur-builds-ai-chief-of-staff-no-tech-background

What keeps Aaron on Genspark is how everything works together in one place. He actively tests other AI platforms and keeps coming back. Speakly alone, he says, is worth the subscription price. He uses it to add voice narration to his daughter's learning world and his coaching profiles, and it's what makes his creations feel alive.

For Aaron, this isn't just about getting more done. His identity expanded. He went from "not a developer" to describing himself as an AI engineer, and he's using that new capability to build richer experiences for his students, his coaching clients, and his five-year-old daughter.

Read the full story → https://www.genspark.ai/blog/how-this-educator-built-apps

I can say I’m the only Nigerian who’s a beta tester for this company. And I see the true potential of scaling if they can come to Nigeria.

We have a lot of population here, especially young people who just finished university or are entering, businesses that still do the manual process, and havent heard of AI that does internal operations, young people to use them and start making Money.

I’ve tried in my best capability to host events through a company, to teach SMEs and real estate companies on how Genspark can make thier work easier and solve problems for them.

But theres more if the team visits Nigeria, hosts a event so that companies can see how this product works and solve thier so many repetitive tasks and slow operations.

For young people, giving them these tools to create their personal brand and businesses, take on gigs and us Genspark to solve problems. This is what my country Nigeria needs.

We need a community or a hub here.

I’d be happy to discuss more, if I’m reached out to

Over a warm cup of tea, we shared our thoughts and experiences in a relaxed and welcoming atmosphere. Together, we explored what is worth preserving amid the constant flow of information, how we can capture it effectively, and how those notes can evolve into meaningful knowledge and action.

A Second Brain is more than a tool for storing large amounts of information. It is another thinking system—one that connects scattered thoughts and inspiration, helping transform them into new ideas and concrete action. Through each other’s questions and experiences, we discovered perspectives that may have remained hidden when working alone.

The Memory Map, which visually represents the relationships between notes and information, seemed to receive an especially positive response from participants. Rather than leaving each note isolated, it allows people to see how individual records connect and expand into a broader knowledge network.

As more notes accumulate, the Memory Map begins to reveal the structure of our thinking and the evolution of our interests—almost like a personal map of knowledge. It can help us understand which topics shape our thinking, identify unexpected relationships between different notes, and uncover new ideas through those connections.

We also had a light conversation about the upcoming Genspark Hackathon next week. Participants shared their interests and early ideas, discussing which topics might be worth exploring and how Genspark could help us approach problems from new perspectives.

Rather than trying to finalize a specific project or outcome, the conversation was about bringing ideas into the open and discovering possibilities together. It was inspiring to see how one person’s question could connect with another person’s experience—and how a vague idea could gradually develop into a clearer direction through conversation.

This meetup was also meaningful because the knowledge accumulated in our Second Brains—and the connections discovered through the Memory Map—could potentially develop into real ideas and prototypes during the hackathon.

Notes become connections, connections become ideas, and ideas grow into collaboration and action.

It was a wonderful gathering filled with warm tea, thoughtful conversations, new connections, and small sparks of inspiration for our next challenge. A heartfelt thank-you to everyone who spent their Sunday afternoon with us.

We hope the thoughts shared during this meetup continue to grow into new questions and ideas within everyone’s notes—and perhaps develop into exciting experiments and outcomes at next week’s Genspark Hackathon.

We look forward to continuing to create opportunities where people can capture, connect, and transform knowledge into real value with Genspark.

See you again at the next meetup—and at the hackathon! ✨

🇺🇸 English

Hello,
I'm Jaehun Go, a Genspark Genius Ambassador.

Since 2015, I have been based on Jeju Island, helping people understand digital tools and generative AI and apply them to their work and learning. Rather than focusing solely on what new technology can do, I focus on what it can help people achieve.

I host a variety of in-person workshops across Jeju, introducing participants to the possibilities of Genspark and practical ways to use it. For those who find it difficult to get started with AI through online resources alone, I create hands-on opportunities to experience Genspark firsthand and explore how it can be applied to their work and everyday lives.

To me, Genspark is a tool that takes ideas beyond simple research and planning and turns them into tangible outcomes, including documents, images, and videos. In my training programs and consulting work, I actively use Genspark throughout the creative process, from storyboarding and producing images and videos to completing the final content.

As a Genspark Genius Ambassador, I am committed to helping more individuals and businesses across Jeju use AI in accessible, practical, and meaningful ways. By sharing proven training practices and insights from my consulting experience, I hope to go beyond simply introducing Genspark's features and help turn its potential into real-world impact and results.

🇰🇷 한국어

안녕하세요.
젠스파크 지니어스 앰배서더 고재훈입니다.

저는 2015년부터 제주도를 기반으로 디지털 도구와 생성형 AI를 누구나 쉽게 이해하고 실제 업무와 학습에 활용할 수 있도록 돕고 있습니다. 새로운 기술의 기능보다 "사람이 이 기술로 무엇을 더 잘할 수 있는가"에 집중하고 있습니다.

제주 지역을 중심으로 다양한 오프라인 워크숍을 진행하며, 사람에게 젠스파크의 가능성과 활용 방법을 알리고 있습니다. 온라인 정보만으로는 AI를 시작하기 어려운 분들과 직접 만나 젠스파크를 경험하고, 자신의 일과 생활에 적용해 볼 수 있는 기회를 만드는 것이 저의 중요한 역할입니다.

저에게 젠스파크는 아이디어를 단순한 검색과 기획에 머무르게 하지 않고, 문서와 이미지, 영상 등 구체적인 결과물로 연결해 주는 도구입니다. 실제 교육과 컨설팅에서도 스토리보드 구성부터 이미지와 영상 제작, 최종 콘텐츠 완성까지 젠스파크를 적극적으로 활용하고 있습니다.

젠스파크 지니어스 앰배서더로서 제주 지역의 더 많은 사람과 기업이 AI를 쉽고 실용적으로 활용할 수 있도록 돕겠습니다. 현장에서 검증한 교육 사례와 컨설팅 경험을 공유하며, 기능을 소개하는 사람을 넘어 젠스파크의 가능성을 실제 변화와 성과로 연결하는 안내자가 되겠습니다.