GoGo | Devpost

Inspiration

GoGo was inspired by leading delivery platforms like DiDi and tech ecosystems like Infosys. We realized that gig-economy couriers often struggle to determine which delivery offers are genuinely lucrative when factoring in distance, traffic, and time. We set out to build an intelligent assistant that brings real-time optimization directly to everyday couriers.

What it does

GoGo is an AI agent that analyzes delivery routes, traffic conditions, and order payouts in real time to optimize courier earnings. By dynamically evaluating incoming delivery requests, GoGo guides couriers toward high-value trips, ensuring they minimize downtime and maximize their hourly earnings.

How we built it

Although we faced a steep learning curve with several unfamiliar technologies, we adopted an iterative rapid-prototyping workflow starting from an intensive brainstorming phase. We built GoGo using modern LLM/AI agent frameworks (OpenAI / LangChain) integrated with location and mapping APIs (such as Google Maps) to compute optimal trade-offs between delivery distance, estimated time, and financial payout.

Challenges we ran into

Our biggest challenge was creating an accurate courier test environment. Modeling real-world variables—such as dynamic traffic, fluctuating payout algorithms, and trip offers—was complex, but essential to rigorously benchmark our AI agent against standard courier decision-making and prove its higher earning efficiency.

Accomplishments that we're proud of

We are proud that our system integrated smoothly across all APIs and that our test simulations proved the AI agent can consistently generate higher earnings compared to traditional, manual courier selection.

What we learned

As our very first hackathon, this was an incredibly rewarding experience. We gained hands-on experience integrating AI agents with external APIs, while mastering rapid problem-solving, teamwork, and communication in a fast-paced environment.

What's next for GoGo

Multi-Platform Aggregation: Expanding support across multiple delivery services (e.g., UberEats, Rappi, DoorDash) into a unified dispatch view.
Predictive Earnings Modeling: Training our model on historical payout data to forecast high-demand surge zones before they happen.
Mobile App & Audio Assistant: Developing a mobile interface featuring hands-free voice guidance for safer, frictionless driving.