Update README.md

README.md

@@ -1,10 +1,110 @@
 ---
-title: README
+title: AI Agent Systems Competition
 emoji: 🏢
 colorFrom: blue
 colorTo: green
 sdk: gradio
-pinned: false
+pinned: true
+sdk_version: 6.5.1
 ---
 
-Edit this `README.md` markdown file to author your organization card.
+# AI Agent Systems Competition (Founding Team Track)
+
+We are running a builder-focused competition to identify exceptional engineers who can design and ship **reliable AI agent systems**.  
+Top performers will be invited to join our founding team / execution center.
+
+---
+
+## 🎯 What to build
+
+Create an **AI agent system** that solves a real business execution workflow.
+
+Choose one track (or propose your own):
+
+### Track A — Sales Follow-Through Agent
+An agent that turns inbound/outbound leads into scheduled calls and updated CRM states (with follow-ups, retries, and escalation).
+
+### Track B — Recruiting Coordination Agent
+An agent that coordinates candidates + interviewers across scheduling, reminders, status updates, and handoff to humans when needed.
+
+### Track C — Open Track (Execution Agent)
+Any execution workflow that has clear ownership + measurable outcomes (support escalation, onboarding, ops tracking, etc.).
+
+---
+
+## ✅ Submission requirements (must have)
+
+Submit **one** entry including:
+
+1) **Code repo (required)**  
+- GitHub or Hugging Face repo link  
+- Must include: `README`, setup steps, and system architecture description
+
+2) **Demo video (required, 3–5 minutes)**  
+Your video must show:
+- A live run of the agent/system
+- A walkthrough of system components
+- At least **one failure case** and how your system handles it
+
+3) **Short write-up (required, 1–2 pages OR README section)**  
+Include:
+- Problem definition & assumptions  
+- Architecture diagram (can be simple)  
+- Reliability & failure handling strategy  
+- Metrics you track (and why)
+
+---
+
+## 🧪 Evaluation rubric
+
+We evaluate systems based on:
+
+- **Reliability & failure handling** (timeouts, retries, fallbacks, human handoff)
+- **Execution logic & ownership** (clear workflow/state, no “prompt-only” demo)
+- **Real-world usability** (deployable mindset, clear interfaces, readable code)
+- **Outcome focus** (metrics, success criteria, measurable impact)
+
+> Flashy UI is not required. We care about systems that can run.
+
+---
+
+## 🗓️ Timeline
+
+- Registration deadline: **[YYYY-MM-DD]**
+- Submission deadline: **[YYYY-MM-DD]**
+- Final demos / interviews: **[YYYY-MM-DD]**
+- Winner announcements: **[YYYY-MM-DD]**
+
+(All times: Pacific Time)
+
+---
+
+## 📮 How to submit
+
+Fill out the submission form:
+
+- **Submission Form:** [PASTE YOUR GOOGLE FORM LINK HERE]
+
+What you will submit in the form:
+- Name / email
+- Track
+- Repo link (GitHub or HF)
+- Demo video link (YouTube / Drive / Loom)
+- Optional: short note (what you’re proud of / what’s incomplete)
+
+---
+
+## 📌 Rules
+
+- Individuals or teams (up to **[N]** people)
+- You may use AI tools during development
+- Submissions must be original work (you can use open-source libraries with proper attribution)
+- We may showcase your demo publicly (only with your permission — specify in the form)
+
+---
+
+## 🙌 Contact
+
+Questions? Post in **Discussions** or email: **[your-email]**
+
+Good luck — we’re excited to see what you build.