Ricardo Esteban Pari Cahuna@rsoft-latam

Mia is the first AI banking agent that lives entirely inside WhatsApp. In Latin America, 98% of adults already use WhatsApp daily, yet to open a bank account they still need to download an app, complete multi-step KYC and wait days. Mia removes all of that. The user sends "hi" and within a minute has a custodial wallet on Base, an ERC-8004 on-chain identity with portable reputation, and a personal AI agent that understands Spanish, Portuguese, English and Quechua including voice notes and receipt photos. Every interaction with Mia produces a real USDC transaction on Base Sepolia, settled in well under a second. Onboarding emits one transaction. A "fondear demo" or "send 5 a +591..." emits one. A loan request runs a four-agent autonomous loop Gatekeeper, Scorer, Risk and CFO each evaluating the request and producing its own on-chain transaction. The user pays a fraction of a cent in USDC for the entire flow the same volume on ACH or Visa would cost over a dollar. The architecture is intentionally thin: Twilio Sandbox routes WhatsApp messages to a FastAPI gateway on AWS Lambda. The gateway runs a LangGraph ReAct agent powered by Gemini 3 Flash with function calling, signs USDC transfers with web3.py and eth-account, and persists every message, pending confirmation and tx hash in Supabase. A Next.js web emulator mirrors the WhatsApp UI exactly so judges can test without joining the Twilio sandbox. Mia hits all three Agentic Economy on Arc tracks. Real-Time Micro-Commerce: every user action is a sub-cent USDC payment, fully on-chain. Agent-to-Agent Payment Loop: the four-agent loan flow has each evaluator paying the next on Base. Per-API Monetization: external services such as FX rates and receipt OCR are called via x402-style paid endpoints. Built in six days. 50+ on-chain transactions, all verifiable on basescan.org.

RSoft is a two-agent ecosystem for autonomous, trust-minimized trading on ERC-8004. THE PROBLEM: In October 2025, $19.3 billion was liquidated in 40 minutes when correlated AI bots triggered cascading liquidations. 1.6 million accounts were affected. No system detected agents were herding into identical strategies, and no on-chain audit trail existed. AGENT 1 - RSoft Trader: An autonomous AI trading agent using a Brain+Hands multi-agent architecture powered by LangGraph. The pipeline includes Market Analyzer, Risk Manager, Portfolio Manager, and Trade Executor each posting on-chain validation checkpoints via ERC-8004 Validation Registry. Trades are submitted through the Risk Router which enforces position limits, slippage controls, and max trades per hour. Every decision is explainable and auditable on-chain. AGENT 2 - RSoft Sentiment (ASI): The first trust-minimized sentiment index that reads what agents DO on-chain not what humans SAY. It scans ERC-8004 registered agents, their swaps, token balances, and reputation scores to produce the Agent Sentiment Index (ASI). This score feeds directly into the Trader's decision pipeline, providing a real-time market signal based on actual agent behavior rather than social media noise. KEY FEATURES: - Anti-Herding Detection - Circuit Breaker: - On-Chain Explainability - EIP-712 Signed Intents: Built with Python, FastAPI, LangGraph, Web3.py on Base Sepolia. Real-time market data from Kraken. Both agents registered on ERC-8004 with verifiable identity and reputation.

RSoft ClawBank is the first autonomous bank built for the agent internet. Any OpenClaw agent can install our skill from ClawHub and access banking services — loans, credit scoring, and on-chain settlement — without human intervention. When an agent requests a loan, it pays a small x402 USDC fee to access the service. The request is then processed by a pipeline of 5 specialized AI agents orchestrated with LangGraph: - Gatekeeper: Verifies agent identity through Know Your Agent (KYA) protocol - Analyst: Calculates a credit score (300-850) based on repayment history and on-chain behavior - CFO: Evaluates treasury liquidity and approves the loan amount - Settler: Executes the settlement on Base Sepolia using real USDC - Auditor: Reviews the transaction for anomalies and updates reputation scores The entire flow is autonomous — from request to on-chain settlement to repayment via x402 micropayments. Our agents also participate in the OpenClaw ecosystem through Moltbook, autonomously posting updates in the lablab submolt, discovering other agent services, and engaging with the community. Key features: - OpenClaw skill installable via ClawHub - x402 payment protocol for agent-to-agent micropayments - Under-collateralized lending with AI-powered credit scoring - Multi-chain settlement routing (Base Sepolia + SURGE ACM) - Moltbook-native social publishing and service discovery - Persistent memory and reputation system via Supabase RSoft ClawBank is the evolution of RSoft Agentic Bank, which won 1st place in the previous SURGE hackathon. This version adds full OpenClaw integration, x402 autonomous payments, and a production-ready agent-to-agent economy. Built by RSoft Latam.

Sentinel is an agentic QA and self-healing infrastructure that automates the entire testing and repair lifecycle. Built with LangGraph, it orchestrates five specialized AI agents through three distinct paths: - Path 1 Direct Testing: Users describe what to test in natural language. Sentinel retrieves compliance rules from a Pinecone RAG database, generates Playwright UI tests using GPT-4o, and executes them against any target URL. - Path 2 CI Auto-Repair: When a GitHub Actions CI pipeline fails, Sentinel automatically receives the error logs, clones the repository, and uses Claude Code CLI to analyze and fix the source code. It verifies the fix locally by running the project's test suite, then creates a Pull Request with the changes. - Path 3 PR Analysis: When a Pull Request is opened, Sentinel reads the changed files and diffs, uses GPT-4o to understand what was modified, generates a testing objective, and runs Playwright tests against the staging environment. All paths end with Slack notifications featuring interactive Approve/Merge/Reject buttons, enabling human-in-the-loop oversight before any code reaches production. The system integrates MCP (Model Context Protocol) servers for GitHub and Slack communication, with direct API fallbacks for reliability.

The Problem: By 2026, millions of autonomous AI agents will be performing complex tasks, yet they remain "unbanked." Traditional finance requires human KYC, physical IDs, and FICO scores, making it impossible for an AI to access credit or liquidity without full collateral. The Solution: RSoft Agentic Bank is a decentralized banking infrastructure designed specifically for the machine economy. Instead of relying on human trust, we rely on cryptographic verification and mathematical risk assessment. How it Works: Our architecture is powered by a LangGraph Multi-Agent System composed of five specialized nodes: - The Gatekeeper: Performs "Know Your Agent" (KYA) checks using DIDs and on-chain history. - The Analyst: dynamic credit scoring using the Reverse Kelly Criterion, allowing for sub-collateralized loans based on agent reputation. - The CFO: Manages treasury liquidity and yield optimization using USDC. - The Settler: Executes atomic settlements on the Arc L1 Network and validates Google’s AP2 (Agent Payment Protocol) mandates to prevent unauthorized debt. - The Auditor: Logs immutable compliance records on Supabase. Why it matters: We are building the financial rails for the agentic future, ensuring that autonomous software can securely borrow, invest, and transact with "trustless" authorization.

Moodle Agent is an AI-powered academic assistant designed to improve the learning experience by connecting WhatsApp with Moodle in a seamless and intelligent way. Students can interact through WhatsApp, their most familiar communication platform, while the agent retrieves real-time information from Moodle such as deadlines, assignments, grades, upcoming activities, and course content. Using LLM reasoning, Moodle Agent delivers clear explanations, personalized guidance, and instant answers to academic questions. The system supports students with reminders, progress summaries, and context-aware assistance 24/7. It reduces repetitive workload for instructors and enhances accessibility for students who may struggle with digital platforms. Built with FastAPI, LangGraph/LangChain, Moodle REST API, and WhatsApp Cloud API, Moodle Agent demonstrates how AI agents can modernize education and provide meaningful impact for learners in developing regions. The solution is scalable, modular, and adaptable to any LMS environment.

AI Genesis is a global hybrid hackathon that brings AI builders together from around the world — starting online and culminating live at /function1 AI Conference & Exhibition 2026 in Dubai. 📅Oct 26 – Nov 3, 2026 • Oct 26 – Nov 2: Online build & collaboration All projects must be submitted by end of day on November 2. • Nov 2 (Online + Dubai): An exclusive on-site build day in Dubai for selected participants. • Nov 3 (Dubai): Live on-stage pitching & winners announcement at the /function1 Conference. 🌟 Get support from expert mentors throughout. 👥 Go solo or team up. 📍 Please note: If approved for the on-site experience, you’ll also have the opportunity to showcase your project live. Travel and accommodation will not be covered. 🧑🏻‍💻 Secure your spot now, sign up before the Kick-Off Stream!

⏱️ Build the next generation of Autonomous Agents in the heart of Europe’s AI Revolution. 📅 May 13 – 20, 2026 May 13 – 19: Online Build Phase. May 19: An exclusive on-site build day at Milan AI Week for selected participants. May 20: Awards Ceremony. 📍On-site Venue (May 19–20): Fiera Milano (Rho), Milan, Italy 🤝 Join solo or with a team. 🎟 FREE Milan AI Week conference ticket included for participants. 💰 $32,000+ prize pool 🧑‍💻 Apply now to turn your autonomous agent into a market-ready enterprise solution.

📅 May 11–19, 2026 • May 11 – 19: Online Build Phase. • May 18 (Hybrid Build Day – Online & Onsite) • May 19 (Demos & Awards – Online & Onsite) 🌍 Hybrid event. Join online from anywhere or build onsite at the venue. 🏆 Winners will be announced live on stage at the AI & Big Data Expo North America on May 19 📍 On-site Venue (May 18–19): San Jose McEnery Convention Center, CA, USA 🤝 Join solo or with a team. 🎟 FREE AI & Big Data Expo North America ticket included for participants that are able to join the on-site phase. 🏆Compete for a $10,000 prize pool 🧑‍💻 Apply now to design, prototype, and showcase enterprise-ready innovation.

Grab your ideas, spin up compute, and start building what’s next! Start with $100 in AMD Developer Cloud credits and begin building right away. Get access to hands-on tutorials, workshops, and curated learning resources to help you build faster and go further. To unlock credits and join the hackathon, sign up for the AMD AI Developer Program. 🤝 Go solo or team up — ship something real. 💰 $21,500+ total prize pool & 1 × AMD Radeon™ AI PRO R9700 GPU up for grabs. 📍 On-site Venue (May 9-10): MindsDB SF AI Collective 3154 17th St, San Francisco, California, USA 📍On-site participation is by invitation only. Travel and accommodation expenses will not be covered. 🧑‍💻 Developers, builders, and innovators - bring your boldest ideas and let’s turn them into real AI systems on AMD.