Contact Memory API for AI Agents

Contact Memory API for AI Agents

Contact Memory API for AI Agents

Our API gives your AI agents persistent memory, enabling them to maintain context, build relationships, and deliver personalized experiences across every interaction.

Our API gives your AI agents persistent memory, enabling them to maintain context, build relationships, and deliver personalized experiences across every interaction.

Our API gives your AI agents persistent memory, enabling them to maintain context, build relationships, and deliver personalized experiences across every interaction.

Deploy Contact Memory in minutes for Agents built on

See API in action

Scenario 1

Attach Documents and their Contents ( PDF, Image ) to Contact Memory

Your AI meets someone new. Without skipping a beat, the Contact Memory API creates a profile, notes key details, and builds context as the conversation unfolds. Even if a name or identity isn’t clear yet, the system can store conversation threads, interests, and preferences for future reference.

Scenario 2

Handles Multiple users with shared contexts across multiple channels

Your AI handles complex multi-participant interactions by design. Each contact maintains its own memory stream while understanding relationships and shared context. One simple API call fetches and updates the right context for everyone in the room.

Scenario 3

Has ultra low latency memory retrieval compatible with voice agents.

Contact Memory API handles sensitive data validation automatically. Every contact lookup first runs through auth gates - DOB match? Email verified? Only after passing the required security checks does your contact memory become accessible. No custom security code needed.

Contact Memory

Help your agents operate in the real world with real people

SOME1

Contact

Memory

Remembers who people are, not just what they said

Builds persistent profiles that evolve over time

Understands organizational hierarchies

Maintains security at the identity level

Cross-references interactions across channels

Progressive learning from each interaction

Understands roles and access levels

Real-time context updates

Maps relationship networks automatically

Identity verification built-in

Ultra Low Latency Memory Retrieval

Handles multi-participant contexts

Time-aware memory management

Auto-enriches contact profiles

Generic

Rag

Memory

Only matches similar text content

Starts fresh with each query

Treats all content equally

Document-level security only

Separate searches for each channel

Static document retrieval

No built-in role awareness

Requires manual document updates

No relationship tracking

No identity management

High Latency Memory Retrieval

Single-document focus

No temporal understanding

Static information only

Use cases

One solution, unlimited use-cases

One solution, unlimited use-cases

One solution, unlimited use-cases

How AI agents build real conversations with real people

How AI agents build real conversations with real people

How AI agents build real conversations with real people

Identity Resolution

Conversation Memory

Security Verification

From Fragments to Person

Turn scattered contact data into a real person - from any email, phone, or account number.

Business Context

Recognize Sarah from Acme Corp across her work and personal contacts.

Smart Contact Matching

Know john.smith@email.com and +1-555-0123 are the same John without asking.

Cross-Platform Identity

Connect website visitor #1234 to support ticket to phone caller - automatically.

Household Recognition

Understand when Mom calls about daughter's account without confusion.

Relationship Mapping

Know this is the accountant who handles five of your client accounts.

Identity Resolution

Conversation Memory

Security Verification

From Fragments to Person

Turn scattered contact data into a real person - from any email, phone, or account number.

Business Context

Recognize Sarah from Acme Corp across her work and personal contacts.

Smart Contact Matching

Know john.smith@email.com and +1-555-0123 are the same John without asking.

Cross-Platform Identity

Connect website visitor #1234 to support ticket to phone caller - automatically.

Household Recognition

Understand when Mom calls about daughter's account without confusion.

Relationship Mapping

Know this is the accountant who handles five of your client accounts.

Implementation

Get started in less than 10 minutes

Get started in less than 10 minutes

Get started in less than 10 minutes

Copy the API implementation code and start your first call.

Copy the API implementation code and start your first call.

Copy the API implementation code and start your first call.

import requests
import json

api_key = 'YOUR_API_KEY'

# 1. Add Contact
contact_name = 'Bob Johnson'
contact_url = 'https://api.some1.com/api/contact/add'
contact_headers = {
    'Content-Type': 'application/json',
    'Authorization': f'Bearer {api_key}'
}
contact_data = {'name': contact_name}

try:
    contact_response = requests.post(contact_url, headers=contact_headers, json=contact_data)
    contact_response.raise_for_status()
    contact_data = contact_response.json()
    print("Contact added successfully!")
    print("Contact Response:", json.dumps(contact_data, indent=2))
    contact_id = contact_data["data"]["contactId"]

    # 2. Add Memory
    memory_text = 'Bob works as a software engineer and likes hiking.'
    memory_tags = ["occupation","hobby"]

    memory_url = 'https://api.some1.com/api/memory/add'
    memory_headers = {
        'Content-Type': 'application/json',
        'Authorization': f'Bearer {api_key}'
    }
    memory_data = {
        'contactId': contact_id,
        'memoryText': memory_text,
         "labels": memory_tags
    }

    memory_response = requests.post(memory_url, headers=memory_headers, json=memory_data)
    memory_response.raise_for_status()
    print("Memory added successfully!")
    print("Memory Response:", json.dumps(memory_response.json(), indent=2))


    # 3. Get Memories of a Contact
    get_memory_url = f'https://api.some1.com/api/memory/{contact_id}'
    get_memory_headers = {
        'Authorization': f'Bearer {api_key}'
    }

    get_memory_response = requests.get(get_memory_url, headers=get_memory_headers)
    get_memory_response.raise_for_status()
    print("Memories fetched successfully!")
    print("Get Memory Response:", json.dumps(get_memory_response.json(), indent=2))

except requests.exceptions.RequestException as e:
    print(f"An error occurred: {e}")

Copy code snippet

import requests
import json

api_key = 'YOUR_API_KEY'

# 1. Add Contact
contact_name = 'Bob Johnson'
contact_url = 'https://api.some1.com/api/contact/add'
contact_headers = {
    'Content-Type': 'application/json',
    'Authorization': f'Bearer {api_key}'
}
contact_data = {'name': contact_name}

try:
    contact_response = requests.post(contact_url, headers=contact_headers, json=contact_data)
    contact_response.raise_for_status()
    contact_data = contact_response.json()
    print("Contact added successfully!")
    print("Contact Response:", json.dumps(contact_data, indent=2))
    contact_id = contact_data["data"]["contactId"]

    # 2. Add Memory
    memory_text = 'Bob works as a software engineer and likes hiking.'
    memory_tags = ["occupation","hobby"]

    memory_url = 'https://api.some1.com/api/memory/add'
    memory_headers = {
        'Content-Type': 'application/json',
        'Authorization': f'Bearer {api_key}'
    }
    memory_data = {
        'contactId': contact_id,
        'memoryText': memory_text,
         "labels": memory_tags
    }

    memory_response = requests.post(memory_url, headers=memory_headers, json=memory_data)
    memory_response.raise_for_status()
    print("Memory added successfully!")
    print("Memory Response:", json.dumps(memory_response.json(), indent=2))


    # 3. Get Memories of a Contact
    get_memory_url = f'https://api.some1.com/api/memory/{contact_id}'
    get_memory_headers = {
        'Authorization': f'Bearer {api_key}'
    }

    get_memory_response = requests.get(get_memory_url, headers=get_memory_headers)
    get_memory_response.raise_for_status()
    print("Memories fetched successfully!")
    print("Get Memory Response:", json.dumps(get_memory_response.json(), indent=2))

except requests.exceptions.RequestException as e:
    print(f"An error occurred: {e}")

Copy code snippet

Request Early Playground Access

Request Early Playground Access

Request Early Playground Access