Bain Capital Ventures

Spencer Penn, Co-Founder + CEO of LightSource, joins Judy Khan Shaw on #NYSEFloorTalk to discuss how AI is revolutionizing global procurement, the biggest trade trends for 2025, and what's next after their latest funding round.

While the spotlight in AI remains fixed on model benchmarks, a quieter - and arguably more consequential - shift is underway. A new standard is emerging that could define how AI systems actually DO things in the real world. Last week, OpenAI committed to MCP. It’s a rare moment of alignment in AI - a shared interface layer gaining traction across rival labs. Everyone from Anthropic (who created it) to OpenAI (its fiercest competitor) is backing it. So what is MCP? Model Context Protocol - is an open standard that lets AI models connect to apps, tools, and data sources in a consistent, composable way.📦 Think: USB-C for AI integrations, or one universal plug that fits everywhere. MCP is the plumbing that turns a model into an agent. It’s not about making models smarter. It’s about making them more usable, agentic, and swappable. It’s the missing layer between intelligence and action. In short: What TCP/IP did for the internet, MCP might do for model-based software. What does it unlock? 🧠For proprietary models like GPT-4 or Claude: Streamlined workflows, plugin ecosystems, and enterprise integrations 🧩 For developers and infra providers: Reusable connectors, less glue code, faster experimentation 🤖 For users: More capable AI agents - ones that don’t just chat, but reason, retrieve, and act What happens next? 👉 Standardization creates surface area. Expect a wave of agent-specific infra, orchestration platforms, and marketplaces - just as Docker and Kubernetes did for containerized workloads. 👉 It shifts the moat. As models become swappable, value flows to the surrounding ecosystem: deployment, governance, interfaces. 👉 It forces composability. Enterprises no longer have to place one big bet. They can mix, match, and evolve their model stack with less friction. If LLMs are the engine, MCP is the wiring. The frontier isn't just smarter models - it's interoperable systems.

I'm hiring a Chief of Staff to work directly with me at FERMÀT - for the right person will set them up to be a business unit leader, or exceptional founder in a few years. If you or anyone you know might be interested, please ping me. Some quick notes on the job: - Reporting to me (CEO), and working with me closely on projects that I currently work on myself. - Must be someone who can ask big Qs and unafraid of highly ambitious goals and experiments. - Will be able to deploy experiments themselves either in the market or within the company and learn *very* fast based on the reactions of the customer. - Can align people to get work done fast. - Comfortable talking to CEOs / executives at any company. What you are likely seeking to do after: - Start your own company - Be the leader of a function / business unit somewhere - Be an early stage VC - Be a COO of a growth stage company Please ping me / DM me directly if you or anyone you know may be interested. Ideally this person will be located in SF (in our SF office), and ok with NY for the right person who is comfortable with fast phone calls to get stuff done.

Charta Health’s CMO and Carbon Health Co-Founder, Caesar Djavaherian, is joining Zahra Shariff (CMO) and Laurie Dotas (COO) of Third Eye Health for a powerhouse discussion on how AI is transforming compliance workflows. 📅 April 2 | 🕘 9–10 am PT 🔗 Save your seat: https://lnkd.in/e7WU89ve They'll be diving into: ✅ How AI reduces compliance risk ✅ The end of painful manual audits ✅ Real-world lessons from early AI adoption If you're in RCM, CDI, compliance, billing, or audits—this one’s for you.

Join us this week for our AI meetup where we’ll break down what it really takes to build and deploy agents that drive real business results. We'll be co-hosting with Mintlify, Prophet Security, and Unstructured at Bain Capital Ventures' offices.

📣 Today 3/31 at 10:10 AM PT, join us for the 8th Advanced LLM Agents MOOC lecture on AlphaProof: When Reinforcement Learning Meets Formal Mathematics by Thomas Hubert Google DeepMind. 🌐 Join the thriving community of the LLM Agents MOOC series, with 23K+ registered learners & ~10K members on Discord! 🚀 Register NOW for the AgentX Competition by BerkeleyRDI University of California, Berkeley, w. sponsors Amazon Hugging Face Lambda Mistral AI Google Groq Schmidt Sciences, and VC partners Accel Bain Capital Ventures Bessemer Venture Partners Lightspeed Mayfield New Enterprise Associates (NEA)! Exciting announcements on prizes/credits/resources and more coming soon! 🙏 Huge thanks to Caiming Xiong Salesforce for the 7th lecture on Multimodal Agents – From Perception to Action. Watch the recording: https://lnkd.in/e-r7qKgQ 🧠 Sign up for the AgentX Competition: https://lnkd.in/ec4AWFbs 📚 Sign up for the Advanced LLM Agents MOOC: https://lnkd.in/gu5fbk_a

Self-made billionaire and serial entrepreneur Marc Lore is far from done. In his latest big bet, Lore has poured over $300 million of his own funds into Wonder, a virtual food hall for the digital generation with ambitions to become so much more. Lore can almost taste the possibility that this will be the one—the company whose name becomes a verb, the killer app that will transform an industry and become his legacy. But first he needs people to understand what Wonder even is. Read the full story in latest issue of Fortune Magazine: https://lnkd.in/eheN2StN

This past week in SF, HSBC Innovation Banking US was thrilled to host some of the top names in the supply chain and logistics space for the first 2025 installment of our Innovation Day event series. The event, produced in partnership with Silicon Foundry , B Capital, Norwest Venture Partners, NewBuild Venture Capital, and Bain Capital Ventures, provided a platform for startups and corporates alike to engage and connect throughout the day around their most pressing questions about innovative manufacturing, procurement and logistics technologies that are driving resilience into supply chains of the future. We’d like to express our gratitude to the stellar C-suite professionals, VCs, subject-matter experts and innovative startups who helped make the event a success. Thank you to all attendees, and a special shout out to our Global Head of Innovation Banking, David Sabow, for delivering closing remarks. HSBC's Marissa Adams Microsoft's Donna Warton Nokia's Nishant Batra Roku's Kevin Callanan B Capital's Gabe Greenbaum Norwest's Sean Jacobsohn Bain Capital Ventures Kevin Zhang NewBuild Venture Capital's Rohini Chakravarthy Silicon Foundry's Neal Hansch and Farzin Shadpour and HSBC Innovation Banking's Alex Choy

Today, Enrique Salem and I are thrilled to announce our seed partnership with Straiker, in collaboration with our friends at Lightspeed. Security is a reactive industry. Hacks happen, security teams rush to find coverage, and risk is managed. The velocity, scale, and ferocity of AI-based attacks renders this model woefully obsolete. The most forward-thinking teams are already planning for a nearby future characterized by human-driven malicious interactions with agents, compromised agents interacting with critical systems, and bad actors using AI to dramatically increase attack intensity. To borrow my partner Saanya Ojha's turn of phrase, there's AI for security, security for AI, and then security from AI. Ankur Shah and Sreenath Kurupati are executing rapidly on all three fronts. We're honored to back this team of deep AI researchers, engineers, and security legends. Personally, I couldn't be happier to be working with friends like Tej Kanani and Girish Chandrasekar. Congrats to everyone on Team Straiker! If you're a forward thinking security leader, or interested in securing against the next generation of breaches, read more here. https://lnkd.in/g7Sb5aCf

Today, Enrique Salem and I are thrilled to announce our seed partnership with Straiker, in collaboration with our friends at Lightspeed. Security is a reactive industry. Hacks happen, security teams rush to find coverage, and risk is managed. The velocity, scale, and ferocity of AI-based attacks renders this model woefully obsolete. The most forward-thinking teams are already planning for a nearby future characterized by human-driven malicious interactions with agents, compromised agents interacting with critical systems, and bad actors using AI to dramatically increase attack intensity. To borrow my partner Saanya Ojha's turn of phrase, there's AI for security, security for AI, and then security from AI. Ankur Shah and Sreenath Kurupati are executing rapidly on all three fronts. We're honored to back this team of deep AI researchers, engineers, and security legends. Personally, I couldn't be happier to be working with friends like Tej Kanani and Girish Chandrasekar. Congrats to everyone on Team Straiker! If you're a forward thinking security leader, or interested in securing against the next generation of breaches, read more here. https://lnkd.in/g7Sb5aCf