Jose Quevedo says he spent an 18-hour hackathon directing AI agents to handle backend work while he concentrated on product flow, visual design, and user experience. His DEV Community post is a brief first-person account, not a build guide: it names FastAPI and MCP servers but does not identify the MVP, show its interface, or explain how its agent worked.
What Quevedo says he did
Quevedo describes himself as a tech architect and UI/UX enthusiast. In his account, he contrasts his approach with writing every line himself and says, “I act as a ‘Tech Visionary’.” He says he directed AI agents to take on backend logic, naming FastAPI and MCP servers, while he focused on product flow, visual design, and user experience. Read Quevedo’s post on DEV Community.
As an Amazon Associate I earn from qualifying purchases.
That division of labor is the core of the story. The post does not give a sequence of tasks, explain how Quevedo reviewed generated code, or say what the AI agents were asked to build. It also does not name the model or service providers involved.
What FastAPI and MCP mean here
FastAPI
FastAPI is a Python framework for building APIs with standard Python type hints. Its documentation shows how to create an API and run a development server, and describes automatic interactive API documentation. That general context explains the framework Quevedo names; it does not verify how it was used in his project. FastAPI.
#1 Best Overall
Model Context Protocol
Model Context Protocol (MCP) is an open protocol for standardizing how applications provide context to large language models. Quevedo mentions MCP servers, but does not say which server he used, what tools or context they provided, or how they behaved. Anthropic’s MCP documentation.
What the post does not establish
Quevedo says the team delivered a “high-performance, secure MVP,” but the post offers no measurements or security evidence to support those descriptions. It does not define what “functional” meant for this MVP or provide a project name, demo, code link, architecture, implementation details, or user-test results. The 18 hours are the author’s description of the hackathon duration, not a productivity benchmark or evidence that the approach will generalize.
Quevedo’s profile lists interests and skills including Figma and several AI coding tools, but it does not connect any of them to this build. The post itself does not say which design or coding tools he used. Quevedo’s DEV Community profile.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →What readers can take from the account
The account offers a useful description of one intended role split: AI agents handled backend work while the human focused on the user-facing experience. It does not provide enough detail to reproduce the project or judge the quality of the generated backend. To assess a similar workflow, a team would need concrete evidence that this post does not supply:
Quick Recap
Best Value
Rank #3
- A clear definition of the user problem and the minimum behavior the MVP must support.
- A description of the backend tasks delegated to agents and how generated code was reviewed and tested.
- A working demonstration of the product flow and agent behavior.
- Evidence for performance and security claims, rather than those labels alone.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
