Personalization Use Cases

Context Handoff

You can now allow your agents, devices, and tools to share the same user state and context — so interactions can continue seamlessly across platforms.

Using Fastino’s Personalization API, your agents can retrieve the user’s latest goals, open tasks, and conversational context, making every transition feel natural and continuous.

Overview

Users expect their assistants to remember context — whether they move from Slack to email, mobile to desktop, or one agent to another.

Fastino provides a unified memory layer that stores conversation history, world model updates, and user state, so context can be transferred, resumed, or shared on demand.

This capability is key for:

  • Multi-agent systems

  • Cross-device AI experiences

  • Tool integrations (e.g., Slack ↔ Notion ↔ Gmail)

  • Enterprise collaboration environments

Key Capabilities

Capability

Description

Example

Seamless Continuity

Resume tasks or reasoning threads between agents or devices.

A Slack message continues into an email draft.

Cross-Agent State Sharing

Share contextual embeddings between specialized agents.

A calendar bot hands off meeting prep context to a notes agent.

Deterministic Context Export

Retrieve structured, portable user state.

Export user focus, goals, and communication tone.

Temporal Context Awareness

Account for when context was last updated.

“This context was last active 2 hours ago.”

Example: Transferring Context Between Tools

Step 1 — Retrieve active context from Fastino

Response

{
  "summary": "Ash is currently preparing a presentation for the Fastino launch. Focused on slide design and messaging refinement."
}

Step 2 — Pass summary to another agent

When the user switches from Notion to Slack, the receiving agent (Slack assistant) initializes with the same summary:

This ensures continuity without repeated user input.

Example: Cross-Device Handoff

When a user moves from desktop to mobile:

  1. Mobile app requests recent context.

  2. Fastino returns portable session memory.

    {
      "summary": "Ash was reviewing the Q4 strategy doc and asked for a summary of market insights."
    }
  3. The mobile agent starts mid-session with full state awareness.

Example: Multi-Agent Context Bridge

Agents can also share context explicitly through Fastino.

For example, a scheduler agent can hand off context to a communication agent:

POST /ingest
{
  "user_id": "usr_42af7c",
  "source": "scheduler_agent",
  "documents": [
    {
      "doc_id": "handoff_meeting_20251027",
      "kind": "handoff_context",
      "title": "Weekly Sync Summary",
      "content": "Meeting scheduled for Thursday 2 PM. Include project roadmap and budget review topics."
    }
  ]
}

Later, the email assistant retrieves this document to draft the appropriate message.

Example: Querying Context for a Specific Tool

POST /query
{
  "user_id": "usr_42af7c",
  "question": "What recent context should I load for the Notion workspace?"
}

Response

{
  "answer": "Ash’s last Notion session involved reviewing the roadmap doc and adding design feedback for the homepage section."
}

This allows tool-specific context rehydration with minimal latency.

Architecture

Context Handoff Flow


Each agent contributes and retrieves context from the same world model, ensuring aligned state across tools.

Example: Deterministic Handoff Snapshot

Response

{
  "summary": "Ash last interacted with the system while preparing a design presentation. The session focused on visuals and stakeholder comments."
}

This can be stored or sent as a JSON payload between agent runtimes.

Integration in Agent Systems

Component

Function

Example

Memory Manager

Writes recent session summaries to Fastino.

/ingest with kind=handoff_context.

Session Loader

Retrieves portable summaries for new contexts.

/summary?purpose=handoff.

Task Coordinator

Passes structured reasoning states.

“What was the last unresolved task?”

Transparency Layer

Explains what context was used.

“Continuing from your last Notion session.”

Example Implementation (Python)

import requests

BASE_URL = "https://api.fastino.ai"
HEADERS = {"Authorization": "x-api-key sk_live_456", "Content-Type": "application/json"}

def save_handoff_context(user_id, title, content):
    payload = {
        "user_id": user_id,
        "source": "notion_agent",
        "documents": [
            {"doc_id": "handoff_doc", "kind": "handoff_context", "title": title, "content": content}
        ]
    }
    return requests.post(f"{BASE_URL}/ingest", json=payload, headers=HEADERS).json()

def get_handoff_summary(user_id):
    r = requests.get(f"{BASE_URL}/summary?user_id={user_id}&purpose=handoff", headers=HEADERS)
    return r.json()["summary"]

# Example usage
save_handoff_context("usr_42af7c", "Roadmap Session", "Ash was reviewing the design roadmap and preparing next steps.")
print(get_handoff_summary("usr_42af7c"))

Use Cases

Use Case

Description

Device Handoff

Continue user sessions between desktop, mobile, and web.

Multi-Agent Continuity

Pass context between specialized assistants.

Cross-Tool Flow

Maintain task awareness across Notion, Gmail, and Slack.

Session Persistence

Save working memory for later resumption.

Transparent AI Collaboration

Let users view what context was shared and when.

Best Practices

  • Use purpose=handoff or purpose=session-handoff for portable summaries.

  • Keep context documents concise (<2KB) to optimize retrieval.

  • Always timestamp handoff events for chronological integrity.

  • Implement transparency: show the user what context is being transferred.

  • Use /delete or TTLs to clear stale session context automatically.

  • Combine with Cross-tool Reasoning for unified coordination across tools.

Example: Session Transparency Log

Agents can record when handoffs occur for user auditability:

POST /ingest
{
  "user_id": "usr_42af7c",
  "source": "handoff_manager",
  "documents": [
    {
      "doc_id": "log_20251027",
      "kind": "handoff_log",
      "title": "Context transferred",
      "content": "Context successfully handed off from Notion agent to Slack assistant at 17:05 UTC."
    }
  ]
}

This reinforces user trust and traceability.

Summary

The Context Handoff use case makes personalization continuous and portable.
By enabling agents to exchange user state and intent through Fastino’s world model, your ecosystem can maintain coherence across sessions, tools, and devices — delivering fluid, memory-aware experiences that feel naturally human.

Next, continue to Personalization Use Cases → Personalized Retrieval to learn how Fastino fetches and ranks user-specific context snippets for adaptive grounding.

On this page