Oversaw the launch of uAgents as the successor to the AEA framework. During this period, the framework grew to over 100K downloads on PyPI and 1.5K stars on GitHub.

Launched agentverse and grew the platform to over 25K users and 2M agents.

Launched DeltaV. The system reached over 10K users during this period before being shut down and succeeded by ASI1.

The wallet went through one major rebranding and one minor rebrand (from Fetch to ASI branding). The product line-up expanded from just a browser-extension for Chrome to include mobile apps (iOS and Android), a browser-extension for Firefox, and later a companion web-app. The companion app included standard web3 wallet features, blockchain explorer features (e.g. block and transaction browsing) as well as unique agent-specific features (e.g. managing agent domain names, agent explorer and more). The Wallet grew by 80% to over 18K active users in that period.

Built Fetch.ai's first end-to-end business analysis system by developing a robust pipeline involving multi-source data collection, data processing, data visualisation, and reporting. The system collects data from various sources (e.g. internal databases, social media channels, token price feeds, analytics platforms such as GA, etc), used Python (Pandas and NumPy) for data processing, Google Cloud Platform (GCP) as the cloud infrastructure, BigQuery for storage and Looker Studio for visualisation, reporting and maintaining a data dashboard.

Created the FIP process to get the public involved in taking ownership of the Fetch.ai blockchain, increasing the decentralisation, transparency and openness of the chain. The process was inspired by Ethereum Improvement Proposals (EIPs) and adapted for the Fetch.ai ecosystem.

Joined the founder to help shape the product and company direction from day one. I advised on vision and strategy, defined the product roadmap, led early design direction, and worked closely with both technical and creative teams to bring the vision to life. Also played a key role in building the initial team while helping lay the foundations for execution and growth.

Oversaw the creation and development of CosmPy.

Directed a team of researchers and developers in delivering the first MVP, leading the design of core components including the underlying economic model, incentive mechanisms, security guarantees, and access control.

Co-created the AEA (Autonomous Economic Agents) Framework along with David Minarsch, Marco Favorito, and others. Within 2 years since launch, the framework was adopted by independent developers, small businesses (e.g. Yoti, Datarella), and large corporations (e.g. Bosch, Festo, ZF group, Catena-X and moveID consortiums) in their respective industries.

Co-created the TAC Framework which allowed running trading agent competitions with various configurations, including a version with a smart contract as the trusted source of truth and competition manager. The framework also provided rudimentary tools to create agents with various strategies to enter the competition. The idea and tools gave way to what later came to be as the AEA framework (see above).

Thesis: The thesis proposes a general computational framework for argumentation-based dialogue that models human-like reasoning with incomplete and contradictory information, supporting enthymemes (natural and imperfect arguments), and diverse interaction modes (e.g., negotiation, persuasion). It utilises a quantitative approach to modelling second-order beliefs via theory of mind in autonomous multi-agent environments.

Project: A tool in Java for explaining argument justification in abstract argumentation framework under multiple semantics using argument game proof theories.