𝗙𝗿𝗼𝗺 𝗗𝗲𝗹𝗶𝘃𝗲𝗿𝘆 𝗥𝗶𝗱𝗲𝗿 𝘁𝗼 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗔𝗜

𝗙𝗿𝗼𝗺 𝗗𝗲𝗹𝗶𝘃𝗲𝗿𝘆 𝗥𝗶𝗱𝗲𝗿 𝘁𝗼 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗔𝗜

I work as a delivery rider from 8 PM to 8 AM.

I spend my spare hours learning to code. In June 2025, I saw a video about AI agents. I wondered if two AI brains could talk to each other to verify facts.

I did not know how to code then. I started learning on May 24, 2026. I have 4 hours every day to study.

Two weeks later, I built my first MVP. It uses several components:

• Multi-expert execution: It calls medical and legal experts at the same time.
• Safety brain: It filters inputs and outputs to keep things safe.
• Memory module: It remembers the last 20 turns of a conversation.
• Director brain: It combines expert answers into one clear response.

The system works by matching user keywords to specific experts. If multiple experts apply, they start a debate. They see the full history of the discussion. This allows them to build on or challenge each other.

It is not perfect. The speed is slow and the output logic is simple. But the core works. I am iterating every day.

I am building this to prove my skills. I want to show my thinking and my judgment.

I have a question for experienced developers:

I follow this weekly schedule:

• Mon/Tue/Wed: Build the MVP.
• Thu/Fri/Sat: Practice Python fundamentals.
• Sun: Review and writing.

Does this rhythm work? How can I learn more effectively?

I am sharing my progress to learn from you. If you have faced similar challenges, please share your story.

Code: https://github.com/ammorick/ai-learning-journey

Full story: https://dev.to/ammorick/from-delivery-rider-to-building-my-first-ai-system-heres-my-story-3d9c

Optional learning community: https://t.me/GyaanSetuAi