AI competitions
Corporate AI hackathon
We design and run intensive hackathons where your own people build an AI prototype inside a fixed window. Vibe coding plus your data and internal LLM turns it into an MVP the business can use straight away.
Overview
A short window, a working prototype
Teams of employees build a working AI prototype inside a fixed window. Vibe coding — describing what you want in plain language and letting AI write the code — has cut build time sharply, which is why these events now pull in planning, design, and business people alongside developers.
Vibe coding
Plain-language instructions cut build time sharply
Your own data
Internal LLM and data make the MVP usable now
Every role
Planning, design, and engineering on one team
How We Run
From picking the theme to shipping the product
Pick the theme
We publish the year's theme, such as agentic AI, along with the internal APIs and data teams can use.
Recruit and form teams
Open to planning, design, and business roles as well as developers.
Prep and environment setup
Onboarding on the vibe coding tools, the internal LLM, and the dev environment.
The main sprint
Teams build a working MVP in a sprint that runs from a few hours to 24.
Judging
MVPs are judged on how finished and how viable they are, sometimes with AI-assisted scoring alongside.
Awards and demo day
The strongest teams present and demo, then take their prizes.
Afterward
We help take the best MVPs to production and spread the practice through the company.
Tracks
Tracks we suggest
We shape the tracks around your work, your data, and what you want out of the event.
Agentic workflow automation
Build AI agents that carry out multi-step tasks on their own.
Internal knowledge RAG chatbot
Build a bot that searches and summarizes internal documents and policies.
Code generation and review
Build a coding assistant that lifts engineering throughput.
Customer support automation
Automate support with call summaries and suggested replies.
Marketing and content generation
Automate copy, images, and campaign production.
Analytics and reporting
Generate BI views and reports from plain-language questions.
Document processing (OCR, summary, translation)
Structure and summarize documents at volume with Document AI.
Security-compliant internal GPT
Build a safe internal GPT wired to an on-premise or private-network LLM.
Judging
A sample scoring sheet
- Technical execution and completeness30 pts
Does the MVP actually run
- Business value and fit with real work25 pts
- Creativity and originality20 pts
- Fit with the theme and problem solving15 pts
- Presentation and demo10 pts
* These weights are a composite of common practice at hackathons in Korea. We adjust them to what the event is for, whether that's collaboration or shipping a product.
Impact
What you get out of it
A stack of working MVPs
A short event produces prototypes that actually run, and a pipeline of candidates for productization.
A development culture that spreads
Hands-on vibe coding raises engineering throughput and changes how teams work together.
Departments that mix
Every role works on one team, and AI literacy rises across all of them.
Faster, safer adoption
Building on internal data and internal LLMs brings secure AI adoption forward.
An in-house AI hackathon, from the brief to the product
Theme design, vibe coding and internal LLM setup, mentoring, judging, prizes, and productization support afterward — we run all of it.