diff --git a/docs/cookbooks/essentials/building-ai-companion.mdx b/docs/cookbooks/essentials/building-ai-companion.mdx
index 95e6df5c9..e6c7e3a57 100644
--- a/docs/cookbooks/essentials/building-ai-companion.mdx
+++ b/docs/cookbooks/essentials/building-ai-companion.mdx
@@ -10,6 +10,32 @@ Problem: LLMs are stateless. GPT doesn't remember conversations. You could stuff
The solution: Mem0. It extracts and stores what matters from conversations, then retrieves it when needed. Your companion remembers user preferences, past events, and history.
+
+
+
+
+ Here we use **Mem0 open source** (`Memory`): all local, no API keys needed for memory. Vectors in **Qdrant**, LLM and embeddings via **Ollama**. The **OpenAI** Python SDK calls Ollama's **OpenAI-compatible** `/v1` endpoint for Ray's chat replies.
+
+ ## Installation
+
+ Install the required dependencies:
+
+ ```bash
+ pip install mem0ai qdrant-client openai ollama
+ ```
+
+ Then start Qdrant and pull the Ollama models:
+
+ ```bash
+ docker run -d -p 6333:6333 qdrant/qdrant
+ ollama pull llama3.1:latest
+ ollama pull nomic-embed-text:latest
+ ```
+
+ You can swap `nomic-embed-text` for any Ollama-supported embedding model (e.g., `snowflake-arctic-embed`, `mxbai-embed-large`). Just update the `model` in the `embedder` config and set `embedding_model_dims` in the Qdrant config to match the model's output dimensions (768 for `nomic-embed-text`).
+
+
+
In this cookbook we'll build a **fitness companion** that:
- Remembers user goals across sessions
@@ -25,6 +51,8 @@ By the end, you'll have a working fitness companion and know how to handle commo
Max wants to train for a marathon. He starts chatting with Ray, an AI running coach.
+
+
```python
from openai import OpenAI
from mem0 import MemoryClient
@@ -53,8 +81,73 @@ def chat(user_input, user_id):
], user_id=user_id)
return response
-
```
+
+
+```python
+from openai import OpenAI
+from mem0 import Memory
+
+OLLAMA_URL = "http://localhost:11434"
+CHAT_MODEL = "llama3.1:latest"
+
+memory = Memory.from_config({
+ "vector_store": {
+ "provider": "qdrant",
+ "config": {
+ "collection_name": "fitness_companion",
+ "host": "localhost",
+ "port": 6333,
+ "embedding_model_dims": 768,
+ },
+ },
+ "llm": {
+ "provider": "ollama",
+ "config": {
+ "model": CHAT_MODEL,
+ "temperature": 0,
+ "max_tokens": 2000,
+ "ollama_base_url": OLLAMA_URL,
+ },
+ },
+ "embedder": {
+ "provider": "ollama",
+ "config": {
+ "model": "nomic-embed-text:latest",
+ "ollama_base_url": OLLAMA_URL,
+ },
+ },
+})
+
+ollama_chat = OpenAI(base_url=f"{OLLAMA_URL}/v1", api_key="ollama")
+
+def chat(user_input, user_id):
+ # Retrieve relevant memories
+ memories = memory.search(user_input, user_id=user_id, limit=5)
+ context = "\n".join(m["memory"] for m in memories["results"])
+
+ # Call LLM with memory context (Ollama via OpenAI-compatible API)
+ response = ollama_chat.chat.completions.create(
+ model=CHAT_MODEL,
+ messages=[
+ {"role": "system", "content": f"You're Ray, a running coach. Memories:\n{context}"},
+ {"role": "user", "content": user_input},
+ ],
+ ).choices[0].message.content
+
+ # Store the exchange
+ memory.add(
+ [
+ {"role": "user", "content": user_input},
+ {"role": "assistant", "content": response},
+ ],
+ user_id=user_id,
+ )
+
+ return response
+```
+
+
**Session 1:**
@@ -62,7 +155,6 @@ def chat(user_input, user_id):
chat("I want to run a marathon in under 4 hours", user_id="max")
# Output: "That's a solid goal. What's your current weekly mileage?"
# Stored in Mem0: "Max wants to run sub-4 marathon"
-
```
**Session 2 (next day, app restarted):**
@@ -70,7 +162,6 @@ chat("I want to run a marathon in under 4 hours", user_id="max")
```python
chat("What should I focus on today?", user_id="max")
# Output: "Based on your sub-4 marathon goal, let's work on building your aerobic base..."
-
```
@@ -87,6 +178,8 @@ Ray remembers. Restart the app, and the goal persists. From here on, we'll focus
Max mentions his knee hurts. That's different from his marathon goal - one is temporary, the other is long-term.
+
+
**Categories vs Metadata:**
- **Categories**: AI-assigned by Mem0 based on content (you can't force them)
@@ -100,7 +193,6 @@ mem0_client.project.update(custom_categories=[
{"constraints": "Injuries, limitations, recovery needs"},
{"preferences": "Training style, surfaces, schedules"}
])
-
```
@@ -121,7 +213,6 @@ mem0_client.add(
[{"role": "user", "content": "My right knee flares up on downhills"}],
user_id="max"
)
-
```
Mem0 reads the content and intelligently picks which categories apply. You define the palette, it handles the tagging.
@@ -135,13 +226,50 @@ mem0_client.add(
user_id="max",
metadata={"workout_type": "speed", "forced_tag": "custom_label"}
)
-
```
+
+
+**Categories via Metadata:**
+
+In open source, model categories with a stable field in `metadata`—here we use `memory_bucket`:
+
+```python
+# Add goal
+memory.add(
+ [{"role": "user", "content": "Sub-4 marathon is my A-race"}],
+ user_id="max",
+ metadata={"memory_bucket": "goals"},
+)
+
+# Add constraint
+memory.add(
+ [{"role": "user", "content": "My right knee flares up on downhills"}],
+ user_id="max",
+ metadata={"memory_bucket": "constraints"},
+)
+```
+
+
+**Categories vs Metadata:** In open source, categories are modeled as `metadata` fields you set on each `add`. Filters only see what you put on `add`.
+
+
+```python
+# Force tag using metadata
+memory.add(
+ [{"role": "user", "content": "Some workout note"}],
+ user_id="max",
+ metadata={"memory_bucket": "goals", "workout_type": "speed", "forced_tag": "custom_label"},
+)
+```
+
+
### Filtering by Category
Retrieve just constraints for workout planning:
+
+
```python
constraints = mem0_client.search(
query="injury concerns",
@@ -155,8 +283,21 @@ constraints = mem0_client.search(
)
print([m["memory"] for m in constraints["results"]])
# Output: ["Max's right knee flares up on downhills"]
-
```
+
+
+```python
+constraints = memory.search(
+ query="injury concerns",
+ user_id="max",
+ filters={"memory_bucket": {"in": ["constraints"]}},
+ threshold=0.0 # optional: widen recall for short phrases
+)
+print([m["memory"] for m in constraints["results"]])
+# Output: ["Max's right knee flares up on downhills"]
+```
+
+
Ray can plan workouts that avoid aggravating Max's knee, without pulling in race goals or other unrelated memories.
@@ -168,12 +309,22 @@ Ray can plan workouts that avoid aggravating Max's knee, without pulling in race
Run the basic loop for a week and check what's stored:
+
+
```python
memories = mem0_client.get_all(filters={"AND": [{"user_id": "max"}]})
print([m["memory"] for m in memories["results"]])
# Output: ["Max wants to run marathon under 4 hours", "hey", "lol ok", "cool thanks", "gtg bye"]
-
```
+
+
+```python
+memories = memory.get_all(user_id="max")
+print([m["memory"] for m in memories["results"]])
+# Output: ["Max wants to run marathon under 4 hours", "hey", "lol ok", "cool thanks", "gtg bye"]
+```
+
+
Without filters, Mem0 stores everything—greetings, filler, and casual chat. This pollutes retrieval: instead of pulling "marathon goal," you get "lol ok." Set custom instructions to keep memory clean.
@@ -183,6 +334,8 @@ Noise. Greetings and filler clutter the memory.
### Custom Instructions
+
+
Tell Mem0 what matters:
```python
@@ -198,11 +351,38 @@ Exclude:
- Casual chatter
- Hypotheticals unless planning related
""")
-
```
+
+
+Tell Mem0 what matters by including `custom_fact_extraction_prompt` in the config dict:
+
+```python
+MEMORY_CONFIG["custom_fact_extraction_prompt"] = """
+Extract from running coach conversations:
+- Training goals and race targets
+- Physical constraints or injuries
+- Training preferences (time of day, surfaces, weather)
+- Progress milestones
+
+Exclude:
+- Greetings and filler
+- Casual chatter
+- Hypotheticals unless planning related
+
+Return JSON with key "facts" as a list of strings (use [] if nothing to store).
+"""
+
+memory = Memory.from_config(MEMORY_CONFIG)
+```
+
+`custom_fact_extraction_prompt` is a top-level key in the config dictionary passed to `Memory.from_config()`. Make sure it's set before creating the Memory instance — not after.
+
+
Now chat again:
+
+
```python
chat("hey how's it going", user_id="max")
chat("I prefer trail running over roads", user_id="max")
@@ -210,8 +390,19 @@ chat("I prefer trail running over roads", user_id="max")
memories = mem0_client.get_all(filters={"AND": [{"user_id": "max"}]})
print([m["memory"] for m in memories["results"]])
# Output: ["Max wants to run marathon under 4 hours", "Max prefers trail running over roads"]
-
```
+
+
+```python
+chat("hey how's it going", user_id="max")
+chat("I prefer trail running over roads", user_id="max")
+
+memories = memory.get_all(user_id="max")
+print([m["memory"] for m in memories["results"]])
+# Output: ["Max wants to run marathon under 4 hours", "Max prefers trail running over roads"]
+```
+
+
**Expected output:** Only 2 memories stored—the marathon goal and trail preference. The greeting "hey how's it going" was filtered out automatically. Custom instructions are working.
@@ -221,8 +412,6 @@ Only meaningful facts. Filler gets dropped automatically.
---
----
-
## Agent Memory for Personality
### Why Agents Need Memory Too
@@ -231,16 +420,30 @@ Max prefers direct feedback, not motivational fluff. Ray needs to remember how t
Store agent personality:
+
+
```python
mem0_client.add(
[{"role": "system", "content": "Max wants direct, data-driven feedback. Skip motivational language."}],
agent_id="ray_coach"
)
-
```
+
+
+```python
+memory.add(
+ [{"role": "user", "content": "Max wants direct, data-driven feedback. Skip motivational language."}],
+ agent_id="ray_coach",
+ infer=False,
+)
+```
+
+
Retrieve agent style alongside user memories:
+
+
```python
# Get coach personality
agent_memories = mem0_client.search("coaching style", agent_id="ray_coach")
@@ -251,8 +454,26 @@ mem0_client.add([
{"role": "user", "content": "How'd my run look today?"},
{"role": "assistant", "content": "Pace was 8:15/mile. Heart rate 152, zone 2."}
], user_id="max", agent_id="ray_coach")
-
```
+
+
+```python
+# Get coach personality
+agent_memories = memory.search("coaching style", agent_id="ray_coach")
+# Output: ["Max wants direct, data-driven feedback. Skip motivational language."]
+
+# Store conversations with agent_id
+memory.add(
+ [
+ {"role": "user", "content": "How'd my run look today?"},
+ {"role": "assistant", "content": "Pace was 8:15/mile. Heart rate 152, zone 2."},
+ ],
+ user_id="max",
+ agent_id="ray_coach",
+)
+```
+
+
**Expected behavior:** Ray's responses are now data-driven and direct. The agent memory stored the coaching style preference, so future responses adapt automatically without Max having to repeat his preference.
@@ -268,6 +489,8 @@ No "Great job!" or "Keep it up!" - just data. Ray adapts to Max's preference.
Don't send every single message to Mem0. Keep recent context in memory, let Mem0 handle the important long-term facts.
+
+
```python
# Store only meaningful exchanges in Mem0
mem0_client.add([
@@ -280,8 +503,27 @@ mem0_client.add([
# "cool thanks" → don't store
# Or rely on custom_instructions to filter automatically
-
```
+
+
+```python
+# Store only meaningful exchanges in Mem0
+memory.add(
+ [
+ {"role": "user", "content": "I want to run a marathon"},
+ {"role": "assistant", "content": "Let's build a training plan"},
+ ],
+ user_id="max",
+)
+
+# Skip storing filler
+# "hey" → don't store
+# "cool thanks" → don't store
+
+# Or rely on custom_fact_extraction_prompt to filter automatically
+```
+
+
Last 10 messages in your app's buffer. Important facts in Mem0. Faster, cheaper, still works.
@@ -293,6 +535,8 @@ Last 10 messages in your app's buffer. Important facts in Mem0. Faster, cheaper,
Max tweaks his ankle. It'll heal in two weeks - the memory should expire too.
+
+
```python
from datetime import datetime, timedelta
@@ -303,10 +547,26 @@ mem0_client.add(
user_id="max",
expiration_date=expiration
)
-
```
In 14 days, this memory disappears automatically. Ray stops asking about the ankle.
+
+
+```python
+from datetime import datetime, timedelta
+
+expiration = (datetime.now() + timedelta(days=14)).strftime("%Y-%m-%d")
+
+memory.add(
+ [{"role": "user", "content": "Rolled my left ankle, needs rest"}],
+ user_id="max",
+ metadata={"memory_bucket": "constraints", "expires_on": expiration},
+)
+```
+
+Store `expires_on` in metadata and prune expired memories in your app. Ray stops asking about the ankle once it's removed.
+
+
---
@@ -314,6 +574,8 @@ In 14 days, this memory disappears automatically. Ray stops asking about the ank
Here's the Mem0 setup combining everything:
+
+
```python
from mem0 import MemoryClient
from datetime import datetime, timedelta
@@ -332,11 +594,55 @@ mem0_client.project.update(
{"name": "preferences", "description": "Training style"}
]
)
-
```
+
+
+```python
+from mem0 import Memory
+from datetime import datetime, timedelta
+
+MEMORY_CONFIG = {
+ "vector_store": {
+ "provider": "qdrant",
+ "config": {
+ "collection_name": "fitness_companion",
+ "host": "localhost",
+ "port": 6333,
+ "embedding_model_dims": 768,
+ },
+ },
+ "llm": {
+ "provider": "ollama",
+ "config": {
+ "model": "llama3.1:latest",
+ "temperature": 0,
+ "max_tokens": 2000,
+ "ollama_base_url": "http://localhost:11434",
+ },
+ },
+ "embedder": {
+ "provider": "ollama",
+ "config": {
+ "model": "nomic-embed-text:latest",
+ "ollama_base_url": "http://localhost:11434",
+ },
+ },
+ "custom_fact_extraction_prompt": """
+ Extract: goals, constraints, preferences, progress
+ Exclude: greetings, filler, casual chat
+ Return JSON with key "facts" as a list of strings.
+ """,
+}
+
+memory = Memory.from_config(MEMORY_CONFIG)
+```
+
+
**Week 1 - Store goals and preferences:**
+
+
```python
mem0_client.add([
{"role": "user", "content": "I want to run a sub-4 marathon"},
@@ -346,11 +652,33 @@ mem0_client.add([
mem0_client.add([
{"role": "user", "content": "I prefer trail running over roads"}
], user_id="max", categories=["preferences"])
-
```
+
+
+```python
+memory.add(
+ [
+ {"role": "user", "content": "I want to run a sub-4 marathon"},
+ {"role": "assistant", "content": "Got it. Let's build a training plan."},
+ ],
+ user_id="max",
+ agent_id="ray",
+ metadata={"memory_bucket": "goals"},
+)
+
+memory.add(
+ [{"role": "user", "content": "I prefer trail running over roads"}],
+ user_id="max",
+ metadata={"memory_bucket": "preferences"},
+)
+```
+
+
**Week 3 - Temporary injury with expiration:**
+
+
```python
expiration = (datetime.now() + timedelta(days=14)).strftime("%Y-%m-%d")
mem0_client.add(
@@ -359,16 +687,36 @@ mem0_client.add(
categories=["constraints"],
expiration_date=expiration
)
-
```
+
+
+```python
+expiration = (datetime.now() + timedelta(days=14)).strftime("%Y-%m-%d")
+memory.add(
+ [{"role": "user", "content": "Rolled ankle, need light workouts"}],
+ user_id="max",
+ metadata={"memory_bucket": "constraints", "expires_on": expiration},
+)
+```
+
+
**Retrieve for context:**
+
+
```python
memories = mem0_client.search("training plan", user_id="max", limit=5)
# Gets: marathon goal, trail preference, ankle injury (if still valid)
-
```
+
+
+```python
+memories = memory.search("training plan", user_id="max", limit=5)
+# Gets: marathon goal, trail preference, ankle injury (if still valid / not pruned)
+```
+
+
Ray remembers goals, preferences, and personality. Handles temporary injuries. Works across sessions.
@@ -380,6 +728,8 @@ Ray remembers goals, preferences, and personality. Handles temporary injuries. W
Training for Boston is different from training for New York. Separate the memory threads:
+
+
```python
mem0_client.add(messages, user_id="max", run_id="boston-2025")
mem0_client.add(messages, user_id="max", run_id="nyc-2025")
@@ -390,8 +740,22 @@ boston_memories = mem0_client.search(
user_id="max",
run_id="boston-2025"
)
-
```
+
+
+```python
+memory.add(messages, user_id="max", run_id="boston-2025")
+memory.add(messages, user_id="max", run_id="nyc-2025")
+
+# Retrieve only Boston memories
+boston_memories = memory.search(
+ "training plan",
+ user_id="max",
+ run_id="boston-2025",
+)
+```
+
+
Each race gets its own episodic boundary. No cross-contamination.
@@ -399,6 +763,8 @@ Each race gets its own episodic boundary. No cross-contamination.
Max has 6 months of training logs to backfill:
+
+
```python
old_logs = [
[{"role": "user", "content": "Completed 20-mile long run"}],
@@ -407,13 +773,27 @@ old_logs = [
for log in old_logs:
mem0_client.add(log, user_id="max")
-
```
+
+
+```python
+old_logs = [
+ [{"role": "user", "content": "Completed 20-mile long run"}],
+ [{"role": "user", "content": "Hit 8:00 pace on tempo run"}],
+]
+
+for log in old_logs:
+ memory.add(log, user_id="max")
+```
+
+
### Handling Contradictions
Max changes his goal from sub-4 to sub-3:45:
+
+
```python
# Find the old memory
memories = mem0_client.get_all(filters={"AND": [{"user_id": "max"}]})
@@ -421,8 +801,19 @@ goal_memory = [m for m in memories["results"] if "sub-4" in m["memory"]][0]
# Update it
mem0_client.update(goal_memory["id"], "Max wants to run sub-3:45 marathon")
-
```
+
+
+```python
+# Find the old memory
+memories = memory.get_all(user_id="max")
+goal_memory = [m for m in memories["results"] if "sub-4" in m["memory"]][0]
+
+# Update it
+memory.update(goal_memory["id"], "Max wants to run sub-3:45 marathon")
+```
+
+
Update instead of creating duplicates.
@@ -430,11 +821,20 @@ Update instead of creating duplicates.
Max works with Ray for running and Jordan for strength training:
+
+
```python
chat("easy run today", user_id="max", agent_id="ray")
chat("leg day workout", user_id="max", agent_id="jordan")
-
```
+
+
+```python
+chat("easy run today", user_id="max", agent_id="ray")
+chat("leg day workout", user_id="max", agent_id="jordan")
+```
+
+
Each coach maintains separate personality memory while sharing user context.
@@ -442,19 +842,44 @@ Each coach maintains separate personality memory while sharing user context.
Prioritize recent training over old data:
+
+
```python
recent = mem0_client.search(
"training progress",
user_id="max",
filters={"created_at": {"gte": "2025-10-01"}}
)
-
```
+
+
+```python
+# Qdrant range filters require numbers — store an epoch timestamp in metadata
+from datetime import datetime
+
+epoch = int(datetime(2025, 10, 15).timestamp())
+memory.add(
+ [{"role": "user", "content": "Completed 18-mile long run"}],
+ user_id="max",
+ metadata={"logged_epoch": epoch},
+)
+
+cutoff = int(datetime(2025, 10, 1).timestamp())
+recent = memory.search(
+ "training progress",
+ user_id="max",
+ filters={"logged_epoch": {"gte": cutoff}},
+)
+```
+
+
### Metadata Tagging
Tag workouts by type:
+
+
```python
mem0_client.add(
[{"role": "user", "content": "10x400m intervals"}],
@@ -468,20 +893,48 @@ speed_sessions = mem0_client.search(
user_id="max",
filters={"metadata": {"workout_type": "speed"}}
)
-
```
+
+
+```python
+memory.add(
+ [{"role": "user", "content": "10x400m intervals"}],
+ user_id="max",
+ metadata={"workout_type": "speed", "intensity": "high"},
+)
+
+# Later, find all speed workouts
+speed_sessions = memory.search(
+ "speed work",
+ user_id="max",
+ filters={"workout_type": "speed"},
+)
+```
+
+
### Pruning Old Memories
Delete irrelevant memories:
+
+
```python
mem0_client.delete(memory_id="mem_xyz")
# Or clear an entire run_id
mem0_client.delete_all(user_id="max", run_id="old-training-cycle")
-
```
+
+
+```python
+memory.delete(memory_id="mem_xyz")
+
+# Or clear an entire run_id
+memory.delete_all(user_id="max", run_id="old-training-cycle")
+```
+
+
---
@@ -515,7 +968,7 @@ Before launching:
- Define 2-3 categories (goals, constraints, preferences)
- Add expiration strategy for time-bound facts
- Implement error handling for API calls
-- Monitor memory quality in Mem0 dashboard
+- Monitor memory quality (Mem0 dashboard or `get_all` / Qdrant when local)
- Clear test data from production project
---