New Post

Author

Rajan Verma
April 09, 2025

AI-Augmented Approach:

a) Feature Scoping & User Stories

  • Provide an LLM (like Claude or GPT-4) with your product vision document and ask it to:

  • Generate comprehensive user personas

  • Draft detailed user stories with acceptance criteria

  • Break down epics into manageable features

  • Identify potential dependencies and edge cases 

Example prompt: "Based on this product overview [paste overview], create a comprehensive set of user stories following the format: 'As a [user type], I want to [action] so that [benefit]'. For each story, include acceptance criteria, potential edge cases, and complexity estimates."

b) Market & Competitor Analysis

  • Feed competitor websites, documentation, and app screenshots into an LLM to:

  • Extract key features and functionality

  • Identify gaps in competitor offerings

  • Suggest potential differentiators 

Example prompt: "Analyze these three competitor product pages [paste URLs/content]. Identify their core features, pricing models, target audience, and apparent strengths/weaknesses. Then suggest 3-5 potential differentiators our product could focus on."

c) Technical Requirements Documentation

  • Use AI to transform high-level requirements into detailed technical specifications:

  • API endpoint definitions

  • Data model drafts

  • Required third-party integrations

  • Performance requirements 

Example prompt: "Convert these user stories [paste stories] into a technical requirements document including: suggested API endpoints with parameters, data models with fields and relationships, third-party integrations needed, and non-functional requirements."

2. Architecture & System Design

Traditional Approach: Whiteboarding sessions, architecture discussions, component diagrams, technology selection.

AI-Augmented Approach:

a) Component & Service Architecture

  • Provide system requirements to an LLM and ask it to:

  • Draft a microservice architecture

  • Suggest service boundaries and responsibilities

  • Recommend communication patterns between services

  • Identify potential bottlenecks 

Example prompt: "Based on these requirements [paste requirements], design a microservice architecture. Include: service boundaries, communication patterns (sync/async), data storage recommendations, and potential scaling considerations."

b) Infrastructure as Code Planning

  • Use AI to generate infrastructure templates and configuration:

  • Draft Terraform/CloudFormation templates

  • Create Kubernetes configuration files