More than 110 developers gathered in Berlin last week for a hackathon focused on building AI-powered payment agents using the Algorand blockchain. The event, organized by the Algorand Foundation, aimed to explore how artificial intelligence can streamline transactions and automate financial services on the network.
The Berlin Hackathon
The hackathon ran over three days, with teams competing to create functional prototypes. Participants worked on AI agents that could handle payments, manage subscriptions, or execute trades autonomously. The winning team built an agent that processes cross-border payments using natural language commands, cutting out the need for manual input.
Organizers said the turnout exceeded expectations, with developers coming from across Europe and beyond. Many were new to Algorand but familiar with AI frameworks. The foundation provided technical mentors and API access to help teams integrate the blockchain's smart contracts and tokenization features.
AI-Powered Payments on Algorand
The projects ranged from simple tipping bots to complex multi-step payment orchestrators. One standout prototype used a large language model to interpret user requests like “send $50 to Alice in euros” and then automatically converted, signed, and broadcast the transaction on Algorand. Another team created an agent that monitors a user's spending and triggers savings transfers based on predefined rules.
Developers said the low transaction fees and fast finality on Algorand made it suitable for AI-driven micropayments. The network’s support for asset creation and atomic swaps also allowed for creative combinations, such as agents that pay for API calls in real time or settle invoice disputes programmatically.
Why Algorand?
Algorand’s pure proof-of-stake consensus and lack of forking appealed to builders who wanted predictable transaction times. The platform’s Layer-1 smart contracts, which run directly on the base layer, simplified the integration with AI models. Several participants noted that the absence of gas bidding wars — unlike Ethereum — made the cost of running an AI agent easier to budget.
The Algorand Foundation has been pushing into the AI and automation space. Earlier this year, it announced a grant program for projects that combine blockchain with machine learning. The Berlin hackathon is part of a broader strategy to attract developers building real-world payment solutions rather than just speculative tokens.
What’s next for the prototypes? Some teams said they plan to continue development outside the hackathon. The foundation hasn’t announced a follow-up event, but organizers hinted at a possible showcase later this year where the most promising agents could be demoed to potential investors and partners.
