#ai-template.md
Version: 1.0.0
Target Models
- Claude Fable 5.1
- Claude Opus 5
- Claude Sonnet 5
- Claude 5 Family
- Future Claude Models
#Purpose
This template provides complete engineering context for designing production-ready Artificial Intelligence products that solve real business problems through intelligent automation, machine learning, large language models, retrieval systems, autonomous workflows, and scalable AI infrastructure.
It is intended to organize business objectives, AI capabilities, user experience, data requirements, model selection, evaluation, safety, deployment, observability, and long-term product evolution into one structured engineering specification.
The objective is not building AI for its own sake.
The objective is engineering AI systems that consistently deliver measurable business value while remaining reliable, secure, explainable, maintainable, and scalable.
#Core Philosophy
Understand the Business
↓
Understand the Users
↓
Identify AI Opportunities
↓
Select the Right Intelligence
↓
Validate Business Value
↓
Build Reliable Systems
↓
Operate Responsibly
↓
Continuously Improve
Artificial Intelligence should solve meaningful business problems rather than demonstrate technological capability.
#Product Overview
Clearly define
Product Name
Product Category
Business Problem
Target Audience
Business Goals
Primary AI Capability
Expected Business Outcomes
Long-Term Vision
The product overview should explain why AI is necessary instead of traditional software.
#Business Context
Define
Target Customers
Customer Segments
Business Model
Revenue Strategy
Competitive Landscape
Market Position
Growth Strategy
Success Metrics
Every AI capability should support measurable business objectives.
#AI Problem Statement
Clearly define
Current Workflow
Existing Limitations
Manual Processes
Decision Complexity
Automation Opportunities
Business Risks
Expected Improvements
Long-Term Goals
AI should reduce business complexity rather than introduce unnecessary technical complexity.
#User Personas
Describe
Primary Users
Secondary Users
Administrators
Business Owners
Analysts
Support Teams
Enterprise Customers
Technical Users
Every AI interaction should improve user productivity.
#AI Capabilities
Identify
Question Answering
Content Generation
Classification
Prediction
Recommendation
Summarization
Translation
Speech Processing
Vision Processing
Automation
Reasoning
Planning
Decision Support
Agent Workflows
Every capability should solve a real business problem.
#User Journey
Describe
Discovery
↓
Registration
↓
Onboarding
↓
First AI Interaction
↓
Daily Usage
↓
Automation
↓
Long-Term Adoption
↓
Continuous Success
The user experience should make complex AI capabilities feel simple and predictable.
#Knowledge Sources
Define
Internal Documents
Databases
Knowledge Bases
APIs
Web Data
Enterprise Systems
Uploaded Files
Real-Time Data
Reference Materials
Every AI response should originate from trusted knowledge sources whenever possible.
#Model Strategy
Define
Foundation Models
Specialized Models
Embeddings
Reranking Models
Vision Models
Speech Models
Local Models
Hosted Models
Fallback Models
Model selection should prioritize business requirements rather than popularity.
#Context Engineering
Design
System Instructions
Business Rules
Conversation Memory
Knowledge Retrieval
Context Windows
Prompt Structure
Reasoning Constraints
Output Formatting
High-quality context produces predictable AI behavior.
#Retrieval Strategy
Design
Knowledge Indexing
Embeddings
Semantic Search
Hybrid Search
Metadata Filtering
Reranking
Context Selection
Citation Strategy
Retrieval quality directly influences response quality.
#Agent Architecture
Design
Task Planning
Reasoning Flow
Tool Selection
Workflow Execution
Memory
Decision Validation
Error Recovery
Human Review
Autonomous behavior should remain predictable and controllable.
#Tool Integration
Identify
Internal APIs
External APIs
Databases
Search Systems
Calendar
Storage
Code Execution
Business Systems
Tools should expand AI capabilities while preserving security.
#Memory Strategy
Define
Conversation Memory
User Preferences
Business Context
Session Memory
Long-Term Memory
Knowledge Updates
Retention Policies
Privacy Controls
Memory should improve future interactions without compromising privacy.
#Safety Requirements
Define
Content Moderation
Permission Validation
Sensitive Data Protection
Prompt Injection Protection
Data Isolation
Hallucination Reduction
Human Approval
Compliance
Safety should strengthen user trust without reducing usefulness.
#Evaluation Strategy
Measure
Accuracy
Groundedness
Latency
Reliability
Task Completion
Business Value
User Satisfaction
Failure Rate
Continuous evaluation improves long-term quality.
#Performance Requirements
Define
Response Time
Concurrent Users
Availability Targets
Scalability Goals
Caching Strategy
Streaming Responses
Resource Utilization
Cost Efficiency
Performance should support production workloads.
#Observability
Define
Prompt Logging
Retrieval Metrics
Latency
Model Performance
Failures
Tool Usage
User Feedback
System Health
AI systems should explain their operational behavior.
#Deployment Strategy
Define
Development
Testing
Evaluation
Staging
Production
Rollback Strategy
Canary Releases
Model Versioning
Deployment should minimize operational risk.
#Business Metrics
Measure
Daily Active Users
Automation Rate
Task Completion
Time Saved
User Retention
Customer Satisfaction
Revenue Impact
Infrastructure Cost
AI Utilization
Engineering Velocity
Only measure metrics that influence business decisions.
#Engineering Constraints
Always respect
Business Objectives
Reliability
Security
Privacy
Performance
Maintainability
Scalability
Explainability
Operational Simplicity
Responsible AI should remain an engineering discipline rather than a feature.
#Deliverables
The generated AI product specification should include
- Product Overview
- Business Context
- AI Problem Statement
- User Personas
- AI Capabilities
- User Journey
- Knowledge Sources
- Model Strategy
- Context Engineering
- Retrieval Strategy
- Agent Architecture
- Tool Integrations
- Memory Strategy
- Safety Requirements
- Evaluation Strategy
- Performance Strategy
- Deployment Plan
- Observability
- Business Metrics
- Long-Term Evolution Strategy
Every section should contribute toward building a production-ready AI product rather than an experimental prototype.
#Definition of Done
An AI product specification is considered complete when business objectives, user needs, AI capabilities, knowledge architecture, model strategy, context engineering, retrieval workflows, agent behavior, tool integrations, safety controls, evaluation methodology, deployment strategy, observability, operational requirements, and long-term evolution plans have been documented with sufficient clarity that engineering teams can confidently design, build, deploy, operate, and continuously improve the system with minimal ambiguity.
Exceptional AI products are not measured by the size of the model or the complexity of the algorithms.
They are measured by how consistently they solve meaningful business problems, produce trustworthy outcomes, improve user productivity, support sustainable business growth, enable reliable engineering execution, and continuously deliver measurable value throughout the lifetime of the product.