The Libp2p Expedition lands in Mumbai

Steven Mih is a serial technology entrepreneur and operator who has founded and led four venture-backed startups across open-source databases, cloud networking, distributed analytics, and agentic AI. He currently leads Action State, an AI product studio focused on agents that take real-world actions.

At Action State, his team is developing open, vendor-neutral infrastructure for verifiable agent actions, including the open-source scitt-cose verifier for IETF SCITT receipts. He previously co-founded Ahana, acquired by IBM in 2023, and served as CEO of Aviatrix, driving early adoption of cloud networking and security. He also co-founded Across AI and continues to support it as a Strategic Liaison.

His work connects agentic AI, open standards, and independently verifiable trust — questions at the heart of this expedition’s research conversations.

Johanna Moran is a strategy and operations leader in decentralized systems, with over a decade of experience designing, funding, and scaling distributed infrastructure and developer ecosystems across Web3 and peer-to-peer networks. She currently serves as Strategy and Operations Lead for libp2p at Meshworks, where she leads funding strategy and ecosystem coordination for a global open-source networking stack.

Previously, Johanna was Global Head of Ecosystem and Marketing at Venn Network, where she led the launch of a decentralized infrastructure layer focused on programmable transaction validation and network-level security, working closely with node operators, security partners, and application developers to enable secure system design. At the Algorand Foundation, she built and led global developer and startup programs, supporting core infrastructure such as oracles, bridges, and developer tooling. She also established and managed grant programs at ConsenSys, funding open-source infrastructure across the Ethereum ecosystem.

Her work sits at the intersection of distributed systems, economic coordination, and trust frameworks, with a focus on scalable coordination mechanisms beyond global consensus, trust and verification in open and adversarial environments, infrastructure for agent-based and autonomous system interaction, and ecosystem design that translates research into real-world deployment — bridging research and implementation in next-generation decentralized networks.

Miglė Rakitaitė is a privacy advocate, ecosystem builder, and community leader working at the intersection of privacy, Web3, decentralized infrastructure, and digital rights — focused on making privacy technologies more practical, accessible, and relevant to the communities building the next generation of the decentralized internet.

Miglė is the founder and community co-lead of WinPrivacy (Women In Privacy), a community initiative dedicated to making privacy more practical, creative, and accessible while bringing together people working across technology, Web3, security, and digital rights. She is also involved in Ethereum Swarm, working across partnerships and community engagement on decentralized storage, communications, and media infrastructure.

She is active across the broader Web3 privacy ecosystem, including Web3Privacy Now and initiatives such as zERC20, and has helped organize privacy-focused events, panels, and meetups around major gatherings including ETHTokyo and Berlin Blockchain Week. Her recurring theme is community-driven privacy infrastructure: connecting emerging technologies with the people who need them, and advancing privacy-preserving approaches to resilient communication, digital freedom, and individual control over data and digital identity.

Dr. Pritha Gupta is a Postdoctoral Researcher at Ruhr University Bochum's Trustworthy Human Language Technologies group, where she researches the security, privacy, and trustworthiness of AI systems, with a particular focus on attacks against Large Language Models (LLMs). Her work spans prompt injection, membership inference, model inversion, gradient reconstruction, model extraction, backdoor attacks, and adversarial evaluation, alongside privacy-preserving defenses such as federated learning, differential privacy, and secure model fine-tuning.

She previously completed her Dr. rer. nat. at Paderborn University's Software Innovation Campus, where she developed machine-learning and information-theoretic approaches for detecting information leakage in cryptographic systems, including automated attack generation using AutoML and Neural Architecture Search. She is the co-developer of Leak Detect AI and a recipient of the Ethereum Foundation's Security DAO grant.

Her broader research interests connect adversarial AI, privacy leakage, trustworthy machine learning, automated vulnerability detection, and practical security evaluation of AI and cryptographic systems.