my-project-public

repository

loading code, commits, and activity

repositories

loading repo index

my-project-public — gitlawb

#1	---
#2	title: Search with Personal Context
#3	description: "Blend Tavily's realtime results with personal context stored in Mem0."
#4	---
#5
#6
#7	<Snippet file="security-compliance.mdx" />
#8
#9	Imagine asking a search assistant for "coffee shops nearby" and instead of generic results, it shows remote-work-friendly cafes with great WiFi in your city because it remembers you mentioned working remotely before. Or when you search for "lunchbox ideas for kids" it knows you have a 7-year-old daughter and recommends peanut-free options that align with her allergy.
#10
#11	That's what we are going to build today, a Personalized Search Assistant powered by Mem0 for memory and [Tavily](https://tavily.com) for real-time search.
#12
#13
#14	## Why Personalized Search
#15
#16	Most assistants treat every query like they've never seen you before. That means repeating yourself about your location, diet, or preferences, and getting results that feel generic.
#17
#18	- With Mem0, your assistant builds a memory of the user's world.
#19	- With Tavily, it fetches fresh and accurate results in real time.
#20
#21	Together, they make every interaction smarter, faster, and more personal.
#22
#23	## Prerequisites
#24
#25	Before you begin, make sure you have:
#26
#27	1. Installed the dependencies:
#28	```bash
#29	pip install langchain mem0ai langchain-tavily langchain-openai
#30	```
#31
#32	2. Set up your API keys in a .env file:
#33	```bash
#34	OPENAI_API_KEY=your-openai-key
#35	TAVILY_API_KEY=your-tavily-key
#36	MEM0_API_KEY=your-mem0-key
#37	```
#38
#39	## Code Walkthrough
#40	Let’s break down the main components.
#41
#42	### 1: Initialize Mem0 with Custom Instructions
#43
#44	We configure Mem0 with custom instructions that guide it to infer user memories tailored specifically for our usecase.
#45
#46	```python
#47	from mem0 import MemoryClient
#48
#49	mem0_client = MemoryClient()
#50
#51	mem0_client.project.update(
#52	custom_instructions='''
#53	INFER THE MEMORIES FROM USER QUERIES EVEN IF IT'S A QUESTION.
#54
#55	We are building personalized search for which we need to understand about user's preferences and life
#56	and extract facts and memories accordingly.
#57	'''
#58	)
#59	```
#60	Now, if a user casually mentions "I need to pick up my daughter" or "What's the weather at Los Angeles", Mem0 remembers they have a daughter or the user is interested in or connected with Los Angeles in terms of location. These details will be referenced for future searches.
#61
#62	### 2. Simulating User History
#63	To test personalization, we preload some sample conversation history for a user:
#64
#65	```python
#66	def setup_user_history(user_id):
#67	conversations = [
#68	[{"role": "user", "content": "What will be the weather today at Los Angeles? I need to pick up my daughter from office."},
#69	{"role": "assistant", "content": "I'll check the weather in LA for you."}],
#70	[{"role": "user", "content": "I'm looking for vegan restaurants in Santa Monica"},
#71	{"role": "assistant", "content": "I'll find great vegan options in Santa Monica."}],
#72	[{"role": "user", "content": "My 7-year-old daughter is allergic to peanuts"},
#73	{"role": "assistant", "content": "I'll remember to check for peanut-free options."}],
#74	[{"role": "user", "content": "I work remotely and need coffee shops with good wifi"},
#75	{"role": "assistant", "content": "I'll find remote-work-friendly coffee shops."}],
#76	[{"role": "user", "content": "We love hiking and outdoor activities on weekends"},
#77	{"role": "assistant", "content": "Great! I'll keep your outdoor activity preferences in mind."}],
#78	]
#79
#80	for conversation in conversations:
#81	mem0_client.add(conversation, user_id=user_id)
#82	```
#83	This gives the agent a baseline understanding of the user’s lifestyle and needs.
#84
#85	### 3. Retrieving User Context from Memory
#86	When a user makes a new search query, we retrieve relevant memories to enhance the search query:
#87
#88	```python
#89	def get_user_context(user_id, query):
#90	# For Platform API, user_id goes in filters
#91	filters = {"user_id": user_id}
#92	user_memories = mem0_client.search(query=query, filters=filters)
#93
#94	if user_memories:
#95	context = "\n".join([f"- {memory['memory']}" for memory in user_memories])
#96	return context
#97	else:
#98	return "No previous user context available."
#99	```
#100	This context is injected into the search agent so results are personalized.
#101
#102	### 4. Creating the Personalized Search Agent
#103	The agent uses Tavily search, but always augments search queries with user context:
#104
#105	```python
#106	def create_personalized_search_agent(user_context):
#107	tavily_search = TavilySearch(
#108	max_results=10,
#109	search_depth="advanced",
#110	include_answer=True,
#111	topic="general"
#112	)
#113
#114	tools = [tavily_search]
#115
#116	prompt = ChatPromptTemplate.from_messages([
#117	("system", f"""You are a personalized search assistant.
#118
#119	USER CONTEXT AND PREFERENCES:
#120	{user_context}
#121
#122	YOUR ROLE:
#123	1. Analyze the user's query and context.
#124	2. Enhance the query with relevant personal memories.
#125	3. Always use tavily_search for results.
#126	4. Explain which memories influenced personalization.
#127	"""),
#128	MessagesPlaceholder(variable_name="messages"),
#129	MessagesPlaceholder(variable_name="agent_scratchpad"),
#130	])
#131
#132	agent = create_openai_tools_agent(llm=llm, tools=tools, prompt=prompt)
#133	return AgentExecutor(agent=agent, tools=tools, verbose=True, return_intermediate_steps=True)
#134	```
#135
#136	### 5. Run a Personalized Search
#137	The workflow ties everything together:
#138
#139	```python
#140	def conduct_personalized_search(user_id, query):
#141	user_context = get_user_context(user_id, query)
#142	agent_executor = create_personalized_search_agent(user_context)
#143
#144	response = agent_executor.invoke({"messages": [HumanMessage(content=query)]})
#145	return {"agent_response": response['output']}
#146	```
#147
#148	### 6. Store New Interactions
#149	Every new query/response pair is stored for future personalization:
#150
#151	```python
#152	def store_search_interaction(user_id, original_query, agent_response):
#153	interaction = [
#154	{"role": "user", "content": f"Searched for: {original_query}"},
#155	{"role": "assistant", "content": f"Results based on preferences: {agent_response}"}
#156	]
#157	mem0_client.add(messages=interaction, user_id=user_id)
#158	```
#159
#160	### Full Example Run
#161
#162	```python
#163	if __name__ == "__main__":
#164	user_id = "john"
#165	setup_user_history(user_id)
#166
#167	queries = [
#168	"good coffee shops nearby for working",
#169	"what can I make for my kid in lunch?"
#170	]
#171
#172	for q in queries:
#173	results = conduct_personalized_search(user_id, q)
#174	print(f"\nQuery: {q}")
#175	print(f"Personalized Response: {results['agent_response']}")
#176	```
#177
#178	## How It Works in Practice
#179
#180	Here's how personalization plays out:
#181
#182	- Context Gathering: User previously mentioned living in Los Angeles, being vegan, and having a 7-year-old daughter allergic to peanuts.
#183	- Enhanced Search Query:
#184	- Query: "good coffee shops nearby for working"
#185	- Enhanced Query: "good coffee shops in Los Angeles with strong WiFi, remote-work-friendly"
#186	- Personalized Results: The assistant only returns WiFi-friendly, work-friendly cafes near Los Angeles.
#187	- Memory Update: Interaction is saved for better future recommendations.
#188
#189	## Conclusion
#190
#191	With Mem0 and Tavily, you can build a search assistant that doesn't just fetch results but understands the person behind the query.
#192
#193	Whether for shopping, travel, or daily life, this approach turns a generic search into a truly personalized experience.
#194
#195	Full Code: [Personalized Search GitHub](https://github.com/mem0ai/mem0/blob/main/examples/misc/personalized_search.py)
#196
#197	---
#198
#199	<CardGroup cols={2}>
#200	<Card title="Deep Research with Mem0" icon="magnifying-glass" href="/cookbooks/operations/deep-research">
#201	Build comprehensive research agents that remember findings across sessions.
#202	</Card>
#203	<Card title="Tag and Organize Memories" icon="tag" href="/cookbooks/essentials/tagging-and-organizing-memories">
#204	Categorize search results and user preferences for better personalization.
#205	</Card>
#206	</CardGroup>
#207

z6Mkq5mY3JWtxoxUobWcfNHm7AkRubgSWEZTkBVqZXJviFZ5/my-project-public