19 Jun 2026Live from Agentforce World Tour London
Joe Inzerillo (Salesforce) on the Fin Acquisition & Agentforce
Joe Inzerillo, President of Enterprise and AI Technology at Salesforce, explains his dual 'customer zero' role running product areas including Slack alongside the office of the CIO. He discusses the just-announced Fin acquisition, contrasting Intercom's polished, SMB-first product and short time-to-value with Agentforce's broader enterprise platform approach, and explains why now was the right time to buy. He says the biggest adoption hurdles are fear and treating agents like set-and-forget software rather than digital labor that needs constant tuning, and warns against trying to agentify an entire human job at once.
Joe Inzerillo on Org Redesign & Rapid Innovation Squads
Joe Inzerillo argues that organizational transformation now matters as much as the technology, using an F1 analogy where people are the speed bumps until they're enabled to become part of the track. He describes Salesforce's rapid innovation squads delivering value in four to six weeks, and Slack shipping roadmap features early thanks to small cross-disciplinary teams. He predicts flatter org charts with more autonomous teams, and says every squad deliverable should include an operations agent so agents end up doing roughly 95% of ops work.
Franny Hsiao (Salesforce): Circular AI & the Second Brain
Franny Hsiao, who leads Salesforce's EMEA team of AI Architects and has worked in AI for over 25 years, traces the field from the Watson/Deep Blue era to today and introduces her 'circular AI' framework inspired by ecology. She describes a personal 'second brain' that flags stale knowledge and feeds team-wide and federated corporate brains, so new joiners contribute learning from day one rather than after months of onboarding. She also argues Europe leads on digital sovereignty and governance-by-design, with customers embracing green, frugal AI by right-sizing models instead of defaulting to the biggest ones.
Franny Hsiao on Frugal AI, Local Models & Human-Centric Design
Franny Hsiao criticizes 'token maxing' as artificial consumption at the climate's expense and champions small, local, open-weight models for tasks like auto-organizing screenshots, plus RAG and batch processing over constant retraining. She predicts a hyper-personalized experience layer where each user works with one chosen agent and the complexity of the wider agent ecosystem is hidden from view. Her core architectural principle for AI-native founders is human-centric design: stalled projects usually fail on psychological safety and resistance to change, not on the technology.
Paul O'Sullivan (Salesforce): London AI Centre & Call to Action
Paul O'Sullivan, SVP and CTO for UK & Ireland at Salesforce, leads a 300-engineer solution engineering team and reports strong momentum for the London AI centre: over 8,500 visitors in 18 months, customers with agents live in production, and AI-literacy sessions for school students. He calls London the world's third-biggest AI hub and frames agentic AI as a once-in-a-lifetime shift playing out vertically across systems of context, work and engagement, and horizontally from outreach through lead-to-cash to after-sales. He says winning customers treat it as a way of life rather than a project — building, iterating and retiring agents as an extension of the workforce — and closes the show with a call to action: start small, scale big, and take your people on the journey.