Wooble
AGENTZERO - Build the Agent, Not the Chatbot
Kavitha’s agency spends hours every Monday manually tracking performance and sending updates. Build MondayAgent, an AI tool that automates analysis, creates simple WhatsApp summaries, and provides a dashboard for quick review and approval.
Evaluationclosed
AIMachine Learning
What you get
Rewards and recognition for participating in this challenge.
Participation certificate
Receive a certificate when you complete the challenge.
Overview
Kavitha’s agency spends hours every Monday manually tracking performance and sending updates. Build MondayAgent, an AI tool that automates analysis, creates simple WhatsApp summaries, and provides a dashboard for quick review and approval.
Problem statement
The Setup
Kavitha runs a 12-person digital marketing agency in Coimbatore. Every Monday, her team lead Rajan spends 2.5 hours pulling Instagram and Google Ads data, updating sheets, calculating weekly changes, writing summaries, and sending 8 WhatsApp updates. Every week. A third of his Monday goes into repetitive work that shouldn’t need a human.
The Problem
Build MondayAgent — an AI agent that automates this entire process. It should take client configs and weekly data, analyze performance, generate simple client-ready WhatsApp summaries, and display them on a dashboard for quick review and approval.
Deliverables
- You must identify one real professional in a reporting-heavy role (account manager, analyst, social media manager, or marketing coordinator). Share summaries generated for at least 3 mock clients with them.
- Collect and submit their feedback covering:
- Would they send these messages as-is?
- What specific changes would they make?
- Does the insight feel human-written or AI-generated?
- This feedback must be included in your final deliverables as proof of real-world validation.
Evaluation criteria
- Insight Quality & Accuracy (data-backed, specific, meaningful) — 30%
- Message Quality (human-like, clear, client-ready WhatsApp summaries) — 25
- System Reliability (handles all 8 clients smoothly, no failures) — 25%
- Real-world Validation (authentic feedback + improvements applied) — 20%