Semantic Memory

Agents need a fact layer separate from conversation logs. Semantic memory answers 'what do I know about X?' without replaying entire chat histories.

Semantic memory is an encyclopedia entry about you — facts without the story of when you mentioned them.

Scenario

Semantic memory holds: {user: 'Alice', stack: 'TypeScript', team_size: 8, preferred_linter: 'eslint'} — retrieved when Alice asks about CI setup.

Solution

In Agent Memory, apply Semantic Memory to this scenario: Semantic memory holds: {user: 'Alice', stack: 'TypeScript', team_size: 8, preferred_linter: 'eslint'} — retrieved when Alice asks about CI setup. Identify the inputs, run the technique, validate the output, and note one thing you would monitor in production.

Do this before moving to the next module — reading alone is not enough.

Open the Code Walkthrough below and run it locally. Change one parameter related to Semantic Memory (e.g. model, temperature, top_k, or tool name), observe the difference in output, and write 2–3 sentences explaining what changed.

Highlighted lines show where Semantic Memory happens in the code.

1# Semantic Memory — minimal example2from openai import OpenAI3
4client = OpenAI()  # create API client5
6# Ask the model to explain this topic7response = client.chat.completions.create(  # core API call for Semantic Memory8    model="gpt-4o-mini",9    messages=[10        {"role": "system", "content": "You explain semantic memory clearly."},11        {"role": "user", "content": f"What is semantic memory?"},12    ],13    temperature=0,14)15print(response.choices[0].message.content)  # show output for debugging

• Treating Semantic Memory as a black box without evaluation
• Ignoring cost and latency in production
• Skipping error handling for semantic memory

Quick recap — the most important points from this module.