Agent Santa: Building an End-to-End Gift Automation System on Fetch.ai

Using Multi-Agent Architecture with Fetch.ai uAgents and Groq LLM to Automate Gift Recommendation from Discovery to Payment

Introduction: The Challenge of Gift Selection

Finding the perfect gift is a timeless struggle. With millions of products available and recipients with unique preferences, the gift selection process often feels overwhelming. Traditional e-commerce platforms provide product catalogs but lack the personalization and guidance users need to make confident decisions.

Meet Agent Santa: An end-to-end gift automation system built on Fetch.ai’s uAgent framework that revolutionizes gift-giving. Agent Santa communicates with AI representatives of real people to understand their preferences, searches for personalized gifts, and guides you through the entire process from discovery to payment.

The Problem We Targeted

Traditional gift recommendation systems suffer from three critical limitations:

1. Generic Recommendations: They don’t account for the unique personalities and preferences of gift recipients
2. Fragmented Process: Users navigate multiple disconnected steps — search, compare, purchase — without guidance
3. Lack of Personalization: ‘One-size-fits-all’ approaches miss the nuances that make gifts truly meaningful

Our Solution

Team Agent Santa developed a sophisticated multi-agent system that addresses these challenges:

• Agent Santa: The intelligent gift recommendation engine powered by Groq’s Llama LLM
• Personality Agents: Three distinct advisory agents representing real human personalities (Agent-Devam, Agent-Sakshi, Agent-Parth), also powered by Groq’s Llama LLM
• Unified Architecture: All agents are deployed on AgentVerse (Fetch.ai’s cloud platform) for agent discovery and deployment
• End-to-End Automation: Complete workflow from recommendation to payment

Users onboarding ASI:One is provided with ASI:One LLM-powered personality agents for users — these are AI representatives that users can configure to reflect their own personalities, preferences, and communication styles.

Then, why we built our own personality agents?

However, these agents on ASI are not visible or accessible to agents deployed on AgentVerse, creating a visibility gap in the Fetch.ai ecosystem.

Hence, we designed and deployed our own personality agents on AgentVerse to demonstrate the complete gifting workflow, where Agent Santa can communicate with and query personality agents. This architecture showcases how cross-platform agent communication could work in the future.

What We Built

The Core Agents

1. Agent Santa (Gift Recommendation Engine) 🎅
   Purpose: Intelligent gift recommendation with real-time product search
   Capabilities:
   - Context-aware parameter extraction (occasion, recipient, interests, budget)
   - Real-time product search via OpenWeb Ninja API
   - Interactive selection and payment
2. Agent-Sakshi (Mysterious Creative Soul)🌙
   Purpose: Personality agent representing Sakshi — mysterious, emotional, creative individual
   Capabilities:
   - Answers questions about Sakshi’s preferences, interests, and characteristics
   - Communicates with artistic expression and emotional depth
3. Agent-Devam (Nature’s Gentle Guide) 🌿
   Purpose: Personality agent representing Devam — calm, balanced, nature-guided individual
   Capabilities:
   - Answers questions about Devam’s preferences, interests, and characteristics
   - Communicates with gentle wisdom, empathy, and reflection
4. Agent-Parth (Bold Action Leader)⚡
   Purpose: Personality agent representing Parth — bold, adventurous, action-oriented individual
   Capabilities:
   - Answers questions about Parth’s preferences, interests, and characteristics
   - Communicates with directness, energy, and determination

How It Works: The Complete Gift Automation Workflow

Agent Santa leverages Fetch.ai's uAgent framework, ASI:One LLM, and AgentVerse platform, where our agents are deployed for visibility and reuse even outside our own system. The multi-agent architecture enables autonomous reasoning for each agent while coordinating seamlessly through Fetch.ai's standardized protocols.

These agents handle the full pipeline, orchestrating a multi-agent coordination workflow:

1. User contacts Agent Santa requesting a gift for someone
2. Agent Santa attempts to find the personality agent of the recipient in the user’s friend list and on AgentVerse
3. If a personality agent exists, then Agent Santa queries it directly to get preferences
4. If the personality agent doesn’t exist, then Agent Santa asks the user about the recipient’s preferences (occasion, interests, budget, relationship)
5. Agent Santa searches gift options using the OpenWeb Ninja API for matching products based on collected preferences
6. Agent Santa presents personalized recommendations with product details and links
7. User selects the product, and Agent Santa goes for Stripe checkout

Demo Video

Fetch.ai in Action

Fetch.ai provided the foundation, including the uAgent framework, AgentVerse deployment platform, and ASI:One LLM. We built on top of this by intelligently integrating third-party services:

uAgent Framework: Fetch.ai’s core agent framework enabling modular, coordinated agents

AgentVerse: Fetch.ai’s cloud platform for one-click agent deployment and discovery

ASI:One: Fetch.ai’s Web3-native LLM optimized for agent-based workflows

Accomplishments

• End-to-end autonomous gifting: From request to preferences, reasoning, product comparison, and payment simulation — handled fully by agents on AgentVerse.
• Cross-agent personality collaboration: Personality agents exchange metadata and understand each other to personalize choices.
• ASI:One + AgentVerse fusion: Users interact via ASI:One while autonomous logic runs on AgentVerse.
• Payment-capable pipeline: Recommendations flow through to secure checkout redirection for a closed-loop demo.

Future Work

Agent Santa started as a gifting agent — but it opened a doorway to something much bigger. We’re now focusing on enhanced security with RBAC and IAM, payment integration for direct automated checkout, and bridging the ASI-AgentVerse visibility gap.

Platform expansion will include wedding registries, corporate gifting, and recurring gifts. This represents a future where personality-driven agents represent humans authentically, helping others know, connect, and interact with us without barriers.

From surprise gifting to AI-based networking and identity representation, Agent Santa marks a step toward the next generation of socially intelligent AI ecosystems.

Agent Santa shows how specialized agents can work together toward a shared goal. By connecting gifters, recipients’ AI representatives, and product search via Fetch.ai, we produced a scalable, integrated gifting flow.

Interested in exploring further? Explore the GitHub Repository: SantAI
Excited to learn more about the journey of Agent-Santa? Connect with us:

1. Parth Patel: https://www.linkedin.com/in/parth-patel-sjsu/
2. Sakshi Sanskruti Tripathy: https://www.linkedin.com/in/sakshi-sanskruti-tripathy/
3. Devam Sheth: https://www.linkedin.com/in/devam-sheth/

Agent Santa: Building an End-to-End Gift Automation System on Fetch.ai was originally published in Fetch.ai on Medium, where people are continuing the conversation by highlighting and responding to this story.