Building AI Startups: Teams, Tech and Takeoff

MBN Solutions, in collaboration with Moonsong Labs, presents an evening focused on the early journeys of AI startups: from initial research ideas to proof-of-concept development, launching a company and what the first few months of an AI startup really look like.

You’ll hear from an Moonsong Labs, an AI incubator, how they combined Web3 and AI, the key skills they needed and where emerging research is heading.

The event showcases two startups at the forefront of innovation: kluster.ai, building decentralised compute infrastructure for scalable AI, and Entourage, developing a shared memory system where AI agents learn from each other. They’ll share how they tackled early challenges, scaling their products and the race to be first to market with a new concept.

Expect practical insights, early lessons, and trend predictions, along with the chance to connect with CTOs, Founders and Senior Leaders building the future of AI.

---------------------------------------------

Mohammed Khaleed (chair)

Principal Consultant Data Science & AI – MBN Solutions

For nearly a decade, Mo has supported startups developing some of today’s most advanced AI systems, ensuring they have the talent needed to drive innovation and growth.

His work is grounded in a strong understanding of the AI and ML landscape, anticipating trends and advising businesses on how to attract and secure the best talent. He specialises in building high-performing Ai/ML teams, from first technical hires through to senior leadership positions including CTOs and Directors of AI and brings a practical, strategic approach to AI recruitment.

https://www.linkedin.com/in/mkhaleed/

------------------------------------------------

Talks / Speakers

Gabriele Farei - Venture Studio Lead, Moonsong Labs

Gabriele leads the AI & Web3 venture-studio arm at Moonsong Labs (MSL), drawing on more than a decade of turning frontier research into enterprise products. Before joining MSL in 2024, he helped launch Corda at R3—now the leading enterprise blockchain—and scaled Qredo’s MPC custody network to $42 billion in processed volume.

A vocal advocate for the coming Agent Economy, Gabriele believes that autonomous AI agents will soon be both the primary consumers and creators of software, reshaping SaaS from static apps into dynamic, self-adapting services.

Twitter – https://x.com/jayfarei

LinkedIn – https://uk.linkedin.com/in/gabrielefarei

Talk summary

In his talk he will unpack MSL’s Agent-Economy thesis, the research and companies the studio has produced in this area, and outline the core skills product teams—across engineering, design, and management—will need to thrive at this inflection point.

-----------------------------------------------------------------

Scott Taylor - Co-Founder & CEO – Entourage

Scott is a serial tech entrepreneur and AI product leader with 15 years’ experience turning frontier technology into scaled businesses. He has:

Built and exited multiple start‑ups, including ed‑tech platform miDrive, and served as Chief Product Officer at cloud‑storage pioneer Livedrive through to its exit.

Led AI product strategy at AIG as Global Head of AI Product, deploying machine‑learning tools across a $320 bn investment portfolio.

Raised more than $100 m in combined venture funding across miDrive and Simba Sleep (former Non‑Exec Director) and was named to Forbes 30 Under 30 (Tech) in 2017.

At Entourage, Scott is creating a shared‑memory layer for agentic AI—technology that captures learning from every execution and makes that collective experience instantly reusable.

LinkedIn – https://www.linkedin.com/in/iamscotttaylor/

Talk summary

In this session Scott will introduce Entourage’s core innovation: a shared, self‑improving memory graph that captures every solved task as executable code, ranks it by real‑world success, and makes that know‑how instantly retrievable by any agent. He’ll explain why this “collective memory” is the missing layer between today’s stateless AI assistants and truly adaptive software, illustrate how the graph powers faster problem‑solving across domains—from ESG reporting to developer onboarding—and outline how companies can pilot the technology to turn their own execution data into a compounding competitive advantage.

----------------------------------------------------------------

Ryan McConville - Head of AI | kluster.AI

Ryan leads AI at kluster.ai, the decentralized deployment network powering scalable, production-ready AI inference and fine-tuning.

With over a decade of experience in AI, he earned his PhD in large scale unsupervised learning, later building and leading an AI research group at the University of Bristol. He has published over 60 papers on topics including distributed and decentralized AI. In industry, Ryan was an ML lead at Weights & Biases before joining kluster.ai as its founding AI scientist.

LinkedIn - https://www.linkedin.com/in/rymcc/

Talk Summary

In this talk, Ryan demystifies the process of building a decentralized AI compute platform and marketplace. He'll dive into key challenges like incentivizing independent GPU providers, maintaining trust and quality for AI consumers, and implementing performance optimization techniques that make cutting-edge LLMs affordable, fast, and ready for real-world production environments

---------------------------------------------------------------

4.00 – 4.30pm – Arrive and welcome by Mohammed Khaleed (chair)

4.30-5.00pm - Gabriele Farei - Venture Studio Lead, Moonsong Labs

5.00-5.30pm - Scott Taylor - Co-Founder & CEO – Entourage

5.30 – 6.00pm – Ryan McConville - Head of AI | kluster.AI

6.00-8.00pm - Networking – food and drinks