MODULE 14 — CONTEXT INJECTION & PROMPT TEMPLATES

6

14.1 What is Context Injection?

Context Injection means adding relevant information to a prompt so that an AI system can produce a more accurate and relevant response.

Without Context

Prompt:

“Explain the policy.”

The AI does not know:

• Which policy?
• Which organisation?
• What specific information is required?

The result may be vague or incorrect.

With Context

Prompt:

“Using the company HR policy document, explain the leave policy for employees.”

Now the AI has specific context and can produce a more useful, document-grounded answer.

14.2 How Context Injection Works

5

Step 1 — User Query

The user submits a question or request.

Step 2 — Retrieve Context

The system collects relevant information such as:

• Documents
• User history
• Company rules
• Examples

Step 3 — Insert Context

The system creates a combined prompt:

User Query + Retrieved Information + System Instructions

Step 4 — Generate Response

The LLM processes the enriched prompt and generates the response.

14.3 Types of Context

5

1. Static Context

Information that remains fixed across interactions.

Examples:

• Company policies
• System instructions
• Tone rules
• Formatting guidelines

2. Dynamic Context

Information that changes according to the current request.

Examples:

• Retrieved documents
• Product information
• Real-time API data
• User preferences

3. Conversational Context

Information from previous messages in the same conversation.

Examples:

• Earlier questions
• Previous responses
• Clarifications
• Dialogue history

14.4 Prompt Templates

A prompt template is a reusable structure for creating consistent AI instructions.

A good template can include:

Role

Defines how the AI should behave.

Example:

“You are a professional teacher.”

Task

Defines what the AI should do.

Example:

“Explain photosynthesis clearly.”

Context

Provides background information or retrieved documents.

Output Format

Defines how the response should be structured.

Example Prompt Template

6

Here:

Role → Expert AI tutor

Task → Explain the topic

Context → {topic}

Output Format → Examples + 3-point summary

14.5 Benefits of Context Injection & Prompt Templates

Benefits

• Consistent AI outputs
• Better control over tone and style
• Improved efficiency
• Easier workflow automation

Best Practices

• Use relevant, high-quality context
• Keep prompts structured
• Keep prompts concise
• Update context regularly

Challenges

Too much context can:

• Confuse the AI
• Increase token usage
• Reduce efficiency
• Exceed context-window limits

Stale or poor-quality context can also lead to inaccurate responses.