PRODUCT REQUIREMENTS DOCUMENT

Project: AI Cost Estimator for GenAI Applications Date: October 2025 Hackathon

SECTION 1: EXECUTIVE SUMMARY / OVERVIEW & OBJECTIVE

This project is a lightweight AI Cost Estimator inspired by appcost.ai, but focused specifically on GenAI-related expenses. By answering five simple questions, users will receive a transparent cost breakdown across model and inference, data, and specialised infrastructure. Objective: Provide a reproducible, accessible tool that empowers stakeholders to make informed decisions about GenAI projects within minutes. Hackathon Alignment: This project not only delivers a working cost estimator, but also practices disciplined cost thinking, strengthens short-cycle engineering practices (PRD → architecture → implementation → metrics → documentation), and demonstrates teamwork under deadlines and expert review.

SECTION 2: SUCCESS METRICS

• Working MVP that runs locally and produces cost estimates
• Setup time: less than 10 minutes from clone to run
• At least three cost categories covered (model and inference, data, infrastructure)
• Transparent assumptions documented in README and ARCHITECTURE.md
• Estimator accepts five baseline inputs (features, usage scale [#users & interactions], interface, requirements, data type/storage)
• Positive feedback from hackathon judges on clarity and usability

SECTION 3: TARGET USERS / PERSONAS

• Engineer Persona: needs breakdowns to plan infrastructure and estimate workloads
• Stakeholder Persona: needs transparent assumptions to evaluate feasibility

SECTION 4: USER SCENARIOS / USE CASES

SCENARIO 1: PRODUCT MANAGER SANITY CHECK

Executive Summary: A product manager wants to quickly estimate the cost of adding a chatbot with summarisation to an internal web app.

Inputs:

• Key Features: Inference-based chatbot
• Target Audience: Internal employees (500 users)