About
I back teams building remarkable software-driven companies. I often partner with them…
Articles by Mike
-
Why I led Lokalise's series A
Why I led Lokalise's series A
I have been studying how digital businesses expand internationally since I first invested in multi-local marketplaces…
13 Comments -
Q and A with Rich Waldron, CEO of tray.io
Q and A with Rich Waldron, CEO of tray.io
When I first met tray CEO Rich Waldron early in 2014, after an introduction from angel (and now VC) Andy McLoughlin…
2 Comments -
Our investment in tray.io
Our investment in tray.io
Imagine an easy way to connect all the software your company runs in the cloud, with the power to customise how data is…
1 Comment -
Fireside chat with Y Combinator President, Sam Altman
Fireside chat with Y Combinator President, Sam Altman
In this podcast, together with my fellow partner Simon Levene we explore with Sam Altman his interest in “hard tech”…
-
When to take Venture Capital
When to take Venture Capital
It may be a surprise to hear this from a venture capitalist, but the vast majority of startups should not take money…
4 Comments -
Welcome Pascal Gauthier, our first Venture Partner
Welcome Pascal Gauthier, our first Venture Partner
The latest post from my partner Toby Coppel We are thrilled to announce that Pascal Gauthier is joining Mosaic as a…
1 Comment -
What funding is right for you?
What funding is right for you?
A big question founders often ask themselves is when should their startup raise venture capital. In the 20 years since…
2 Comments -
Our investment in Blockstream
Our investment in Blockstream
The latest post from my partner Simon Levene Mosaic recently participated in the $55M Series A financing for…
5 Comments -
Body and soul
Body and soul
One of my partners was heading past me the other day with an entrepreneur and as he ushered her in, he asked "are you…
Activity
10K followers
-
Mike Chalfen shared thisExcited to see this launch of Knowlix and its all too rare approach that truly thinks through agentic operations fully from the ground up. Less dopamine, more controlled power and sustained and ever-increasing coordination. Well done Francesco Wiedemann and the mighty Peter Georg Meier and team.Mike Chalfen shared thisIntroducing Knowlix - AI work platform for small business. Here is a secret big companies have known for years. They run everything - sales, customers, projects, employees - on one connected platform. It's called Enterprise Resource Planning (ERP). It allows them to have a scalable back office, instead of the chaos many small businesses live in. But there's a catch. A normal ERP takes ~9 months and an army of consultants to set up. Almost no small business can afford that. So most are left with three options: ⒈ Run on 15+ disconnected apps. ⒉ Try your best at becoming a vibe coder to build your own apps from scratch. ⒊ Connect your disconnected apps and hope the AI agent on top doesn't go rogue. So we've decided to create a 4th option: an AI-native ERP, affordable for businesses of any size. ➤ Secure work platform that tailors itself to your business, in minutes instead of months. ➤ An AI teammate does the busywork for you within clear guardrails. So you can fall back in love with your business 🧡 Comment “KNOWLIX” & we'll give you extra credits!
-
Mike Chalfen reposted thisMike Chalfen reposted thisWe are growing faster than at any point in Levity's history, and we are hiring across the board to keep up. Logistics runs on email. The whole industry coordinates trillions of dollars of work through inboxes: quotes, orders, exceptions, invoices. We built the AI that finally reads it, structures it, and runs it, with people in control. We call it Communication Intelligence for the global supply chain. And it is working. Companies like Gebrüder Weiss, LKW WALTER, ECU Worldwide, Ascent Global Logistics, and DPD have stopped asking whether AI can do this and started asking how fast we can roll it out. Demand is pulling us forward faster than we planned. So we are growing every team: engineering, go-to-market, customer delivery. We are a global, remote-first team headquartered in Europe, so we hire the best people wherever they are and build the company around them. If you want to build AI that runs real operations for the world's biggest logistics companies, not another demo that breaks in week two, this is the place. And if you know someone brilliant, send them our way. We pay $5,000 for every hire who comes through a referral, and I mean it. We are out to run the world's logistics on AI. That is a bigger job than today's team can finish, which is the best reason I know to join now.
-
Mike Chalfen reposted thisMike Chalfen reposted this"We want to see more innovation of that kind, opening new routes to market and supporting responsible AI development." That was the Rt Hon Rachel Reeves at the AI Adoption Summit this week, calling out the UK's "Harbr-powered" Creative Content Exchange 👏 In a speech outlining her bold ambitions to make the UK the fastest AI adopter in the G7, the Chancellor stressed that success will depend on a few specific conditions: access to data, clear regulation, the ability to commercialise, and world-class infrastructure. Projects like the Creative Content Exchange are seeking to enable exactly that. By exploring how creative content can be shared securely, discovered more easily, and accessed on clear commercial terms, the initiative is helping unlock new opportunities for home-grown innovation – while offering fair value for British creators. As the UK's AI strategy continues to take shape, we're proud to be playing a small part in its progress. You can read the whole speech here: https://lnkd.in/eHstFytv
-
Mike Chalfen reposted thisMike Chalfen reposted thisEvery new technology has a dumb first phase. We bolt the new thing onto the old system. AI in finance and consulting is mostly in that phase. The history books will laugh at us. The future belongs to systems that are entirely redesigned around AI. Two questions worth thinking about: AI generates faster than people can review. What happens when that gap keeps widening? If every firm is running the same models on the same tools, what stops you from becoming a thinning wrapper around an API? Why can't your clients go straight to Claude? If you want to know what that next phase looks like, you can read my article here: https://lnkd.in/eQ72pBAEGun Turrets on a Sailing Ship: moving past the dumb first phase of Enterprise AIGun Turrets on a Sailing Ship: moving past the dumb first phase of Enterprise AI
-
Mike Chalfen shared thisBIG news for entrepreneurs and our ecosystem, and a reflection of a killer combination of a massive market opportunity x brilliant founders x grit. I am delighted to have supported and worked with Entrepreneurs First and Alice Bentinck and Matt Clifford since 2016, and can only echo Alice in calling out the amazing Tom Shinner for his contribution too.Mike Chalfen shared thisToday, we're announcing Entrepreneurs First's Series D: $200M of fresh capital, at a unicorn+ valuation. We're in a golden age of entrepreneurship. More individuals than ever before are taking the leap, with the world's most transformative technology at their fingertips. For over a decade, we've backed extraordinary people without asking what idea they're working on. We call this Talent Investing; believing in someone from Day 1 and giving them the peer group, environment, and capital to find their life's work. And it works! Companies built through EF are now worth $16bn, up from $3bn at our last raise in 2021. Thank you to our investors for their continued belief, the EF team for bringing this vision to life, and the founders who rejected the status quo to build globally important companies with us. And a huge thank you to Matt Clifford, my exceptional cofounder, and Tom Shinner for being instrumental in making this happen.
-
Mike Chalfen reposted thisMike Chalfen reposted thisEveryone's talking about which LLM is best for agents. But after reading Lance Martin's deep dive into Manus's architecture, I'm convinced the real differentiator isn't the model. It's the context engineering around it. The problem: A typical Manus task makes ~50 tool calls. Without careful management, results pile up in the context window and the LLM gets dumber the longer it runs. Anthropic calls it "attention budget depletion." Chroma calls it "context rot." Manus solves this with three strategies: 𝟏. 𝐑𝐞𝐝𝐮𝐜𝐞: Older tool results get compacted to file path references. When that's not enough, the full trajectory gets summarized via structured schema. 𝟐. 𝐎𝐟𝐟𝐥𝐨𝐚𝐝: Fewer than 20 atomic tools bound to the LLM. Everything else runs in a sandbox via bash. The model stays lean. 𝟑. 𝐈𝐬𝐨𝐥𝐚𝐭𝐞: Sub-agents exist to give each task a clean context window, not to mimic human roles. No context bleed. On top of that, Manus doesn't commit to one model. It routes tasks; Claude for code, Gemini for multimodal, OpenAI for math. Model-agnostic by design. The takeaway: Manus has been refactored 5 times since launch. If a stronger model doesn't improve your agent, your engineering is the bottleneck. Stop obsessing over which model to pick. Start obsessing over the implementation around it. Full article: https://lnkd.in/gaY3rAV5
-
Mike Chalfen shared thisFantastic to see this level of enthusiasm from Seb Johnson 📊🇪🇺 in support of the incredibly ambitious SurrealDB team. With this release, Surreal aims big: providing the definitive agent-ready database. I am also a proud supporter, and delighted to have invested in this funding round.Mike Chalfen shared thisBREAKING: SurrealDB has raised a $23m Series A boost AND launched SurrealDB 3.0! LETS GO The UK startup, founded by brothers Jaime Morgan Hitchcock and Tobie Morgan Hitchcock, is a cloud-native, multi-model database built to power the AI-native applications. It raised a Series A back in 2024 from FirstMark (Matt Turck) and Georgian (Emily Walsh), and now announced an additional $23m from Chalfen Ventures and Begin Capital. As a result Mike Chalfen join the company as a Director. The company is also releasing SurrealDB 3.0 - a new database designed to unify multiple data models into a single solution allowing developers to build AI applications faster and more efficiently. This is VERY COOL. I love seeing UK and European startups building at the infrastructure layer that is enabling AI companies to take off. I interviewed both Jaime and Tobie on the Scaling Europe show where we discussed: 🚀 How SurrealDB became the fastest-growing db of all time 🚀 The new round and what it will unlock 🚀 Sales, GTM and US expansion Check it out by searching "Scaling Europe - Tobie and Jaime" on Youtube or Spotify!
-
Mike Chalfen shared thisExcellent coverage of the epic SurrealDB 3.0 release by Sean M. Kerner in VentureBeat. https://lnkd.in/e5_gxqBn The article has real substance and is worth reading carefully. Every compute era requires a new database paradigm. So far in the AI era, most ambitious AI projects stall in enterprises. They need a data platform that makes unprecedentedly large scale contextual information available to agentic systems, in a way that is synchronised across data sources, fast, and secure. SurrealDB is that platform. I am proud to invest in the company and to have been chosen as their new partner by founders Tobie Morgan Hitchcock and Jaime Morgan Hitchcock.SurrealDB 3.0 wants to replace your five-database RAG stack with oneSurrealDB 3.0 wants to replace your five-database RAG stack with one
-
Mike Chalfen shared thisWhat an incredible product announcement. I am proud to have invested in SurrealDB and backed the remarkable Tobie Morgan Hitchcock and Jaime Morgan Hitchcock. Every compute era requires a new database paradigm. So far in the AI era, too many ambitious AI projects stall in enterprises. Why? They need a data platform that makes unprecedentedly large scale contextual information available to agentic systems, in a way that is synchronised across data sources, fast, and secure. SurrealDB is that platform. And since it works a whole system - with agent frameworks, protocols, tools, integrations & native agent memory management - it can be used effortlessly. It meets the needs of AI agents and enterprise data governance. It is the best on-ramp for companies looking to get native AI initiatives off the ground, and I believe that it can shape what it means for a business to be agent-ready.Mike Chalfen shared thisToday marks an important moment for us. SurrealDB 3.0 is now generally available. It is our most stable, performant, and enterprise-ready release to date, and a meaningful step forward in how AI systems manage memory and context. AI agents struggle with durable state. They forget facts, lose relationships, and fragment context as complexity grows. SurrealDB 3.0 addresses that directly by enabling first-class agent memory and context graphs inside the database itself, so models run next to the data, context remains synchronised, and application logic becomes simpler and more dependable. Alongside the release, we are announcing our $38M Series A, bringing total funding to $44M including seed. We are sharing this now because the capital and the product go hand in hand. It allows us to support 3.0 at scale and continue investing in reliability, performance, security, cloud, and enterprise readiness. To our current and new investors - FirstMark, Georgian, Crew Capital, Alumni Ventures, Chalfen Ventures, and Begin Capital - thank you for your continued belief and partnership. To our brilliant team, this reflects the focus and standards you bring to the work every day. To our community, your feedback and belief have shaped this release more than you know. If you are building agentic systems, I would encourage you to take a serious look at SurrealDB 3.0. This release lays the foundation for what comes next.
Mike Chalfen commented on a post
3w
Adventure sound just right for you Luca Bocchio - as you once said to me, “the important question is: what can go right?”
Mike Chalfen replied to a comment
1mo
Helen Murphy v kind indeed , and would write your first cheque again too!
-
Mike Chalfen liked thisMike Chalfen liked thisToday Isembard is launching the largest factory to open in London since WW2. Southwark will be our new global headquarters, where we hire and train the vanguard of a new industrial age. Manufacturing engineers, software engineers and AI researchers will all coalesce around a real factory producing at scale the components needed to rebuild a new kind of industrial base across the United Kingdom and the rest of the world. The location is also very near to where Brunel spent his early career driving the Thames Tunnel out from Rotherhithe. I think its quite an apt allegory. A seemingly hopeless endeavour that is new and whose critics are many. But for those who believe, it seems obvious. If you cannot shake the feeling your life is called for greater things, come speak to us. We need all the help we can get it. Small acorns, mighty oaks.
-
Mike Chalfen liked thisMike Chalfen liked thisDashboards are dead. Long live Polar Operator. Most AIs are strangers to your business. They are like contractors. You send them data. Explain what it means. Double-check the work. They live in a parallel stack that has no idea what "net revenue" means for you. So they guess. And hallucinate. We shipped the other thing. Today, we’re launching Polar Operator: The in-house AI hire that lives on top of your data warehouse and semantic layer. It uses your definitions. It knows your channels. Your margins. Your inventory. The numbers your team already trusts. It runs the recurring work in Slack. It acts when you approve. What it has already done for brands: → Caught $3k/day of Meta spend hitting a broken URL in its daily landing page audit. → Matched live inventory against active campaigns and caught $7k spent on out-of-stock products. → Built an hourly Black Friday tracker across ads, inventory, and sales -> Diagnosed soft email performance and drafted on-brand Klaviyo campaigns → Flagged a manufacturing defect from shipment quality data Nothing to host. No keys to manage. A Polar engineer in your channel while your team trains it. The best agent is not the one that knows everything. It is the one that knows your business. Polar Analytics 🐻❄️ helped commerce teams trust their numbers. Polar Operator puts that to work.
-
Mike Chalfen liked thisMike Chalfen liked this"Consultancies are done. Partners are leaving because they think their firms are a zero." I've heard this enough times now from Silicon Valley insiders that I wrote a blog about it. I do not agree that consultancy is dead. But the Silicon Valley people are right about one thing: the models will be smarter than us. Long-term, the chances of human intelligence adding real value over models are, as the Bay Area crowd would say, "zero". If your selling point was being "the smartest person in the room", you're in for a rough time. Luckily, you've got options. Consultants are in the problem-solving business, and intelligence alone does not solve most problems. Intelligence needs sidekicks. These are its complements. Consultants sell at least three powerful complements to intelligence: → Proprietary Information. A client is seeing a problem for the first time; the consultancies have seen it a thousand times. Today, that asymmetry is packaged as experienced partners. In the future, it will be proprietary data for AI. → The Squishy Stuff. Business problems are just as often social as intellectual. You don't solve this with more IQ. You solve it with listening, story-telling, incentivising, deal-making, reorganising — all uniquely human undertakings. → Independence (and liability). A firm is not allowed to audit itself. Even if the superintelligent AI is neutral, someone independent needs to write its prompts. Economically speaking, if intelligence is powerful but commoditised ("intelligence too cheap to meter"), the very best thing you can be is its complement. As demand for the commodity explodes, so does demand for your complement — except you can charge a higher margin. Of the three complements, proprietary information is the one that compounds. The models have read everything ever published about post-merger integrations. But they have no idea what happened on the inside; the consultants do. As cognition shifts from humans to AI, the most urgent task for every consultancy is to rebuild this learning loop in digital form. It's a true fork in the road. The consultancies that build the right operating system for AI will entrench themselves further with each new project. The ones that fail will quickly be outcompeted. Full piece here: https://lnkd.in/eCSiAVNw
-
Mike Chalfen liked thisMike Chalfen liked thisToday is a very big day for HiBob. Salesforce is leading a $166 million investment in HiBob alongside Farallon Capital Management. I’m incredibly proud of what this moment represents, and incredibly grateful to the village that got us here. To our Bobbers, thank you for the ambition, creativity, and determination you bring every day. To our customers, partners, and investors, thank you for believing in HiBob, challenging us, and helping us build something that matters. We’ve always believed that workforce data can become something much bigger than HR data. As AI changes how organizations operate, trusted people context will become increasingly important to how businesses make decisions and how people and AI work together. That is the vision we’re building toward with Organizational Intelligence: helping organizations turn workforce data and trusted people context into a deeper understanding of how their business works, so they can make better decisions and act with greater confidence. This investment is a huge signal that we’re on the right path, and gives us even more fuel to accelerate our product, technology, AI, and innovation roadmap. A very big moment for our village. And a very exciting next chapter. More on what this means from me here: https://lnkd.in/ejAEQF7q
-
Mike Chalfen liked thisMike Chalfen liked this👇 What's going on there? One person is our dynamic AI Minister, Kanishka Narayan MP. The other is Jayne - an heroic social worker from Vale of Glamorgan Council. Jayne's severely hearing impaired and lipreads during her assessments. Today, the council relies on Beam technologies to transform the efficiency, but above all the human experience. Jayne no longer has to fear missing things. People in meetings no longer sit waiting as she types. They no longer are made to feel - through no fault of Jayne's - like they're being processed through a system. Just human connection - as it should be, in social services. "I rely on lipreading during my assessments, so Notes has changed things immensely for me. The biggest difference is that patients can look at my face rather than the top of my head." For me, that picture sums up the #AI story the British people want: British businesses improving services for British people. Thank you Kanishka Narayan MP and your team for shining a spotlight on work that matters - above all that of the incredible team at Vale of Glamorgan Council. #socialimpact #socialservices #techforgood
-
Mike Chalfen liked thisMike Chalfen liked thisClosing breakfast with the Bridge Labs leadership team this morning. Daniel Guevara, you were missed. Ken Takagi, Alessandro Romano, Sergio Gámez and Alex Crooks: thank you. It has been a pleasure to welcome you and your wider team into the business and to feel your passion and impact already. Worth saying something about how this team got here. Bridge Labs did not build a business by raising money. They built one by serving Sedna customers on our open platform. They found real problems in commodity trading, chartering and operations, built for them, and got paid directly for the value they delivered. Bootstrapped. Useful from day one. That path is still open. Chartering, operations and trade execution is not one market. It is tens of thousands of businesses: owners, charterers, brokers, operators, ship agents, traders and suppliers. Different fleets, cargoes, trades, jurisdictions and ways of working. That variety is exactly why no single product will ever cover it, and why the application opportunity is so large. I cannot buy every company that builds on Sedna. That is not the point. The point is that you can build a real business on our platform, bootstrapped or venture backed, and deliver real value to our users, our customers and the wider ecosystem. You can make real money. Shipping underwrites global trade. Not enough good software serves it. If you want to build for impact in chartering, operations and commodity trade execution, reach out to me or to partnerships@sedna.com. #Maritime #Shipping #Chartering #MaritimeTech #Partnerships #AI
-
Mike Chalfen liked thisMike Chalfen liked thisYou can now email Tracelight. "Implement all my client's comments on this PowerPoint deck" "Update this model with new assumptions and explore scenarios with an app" Send it, get back to your other tasks. Tracelight works in the background, and the finished work lands in your inbox in minutes. Try it now at https://tracelight.ai
-
Mike Chalfen liked thisMike Chalfen liked thisFOCUS IS SUDDENLY A DISADVANTAGE "The man who does not read has no advantage over the man who cannot read." I can't improve on this quote that implores people to read books, but I can borrow its shape. "The startup that does not use its agility has no advantage over the incumbent that has none." That has always been true. What's changed in the AI era is where agility shows up. It used to be strategic: pivot faster, ship faster, reach the customer before the incumbent noticed. Now it's structural. Development velocity has collapsed the cost of building a second product, and a third, and the customer wants everything to become agentic. This produces a harder version of the same sentence: "The AI startup that does not use its agentic development velocity to build a broad product portfolio has no advantage over the SaaS incumbent with three thousand engineers." This runs against two decades of dogma. Focus was the first commandment of SaaS, and for good reason. A narrow product meant a better product, a clear ICP, and a go-to-market aimed at one knowledgeable buyer inside the organization who was empowered to sign. During the SaaS boom, purchasing power spread out to individual teams and department managers. In the AI era that has changed. Focus is an initial wedge, not a strategy. Stay narrow too long and you cap yourself at the buyer who owns a workflow, and that buyer cannot authorize the organizational change that moves a real line on the P&L. Only the people above them can, and they aren't buying software. They're buying automated workflows and a new org chart. Which raises the obvious question. If everyone can build everything, what is left to compete on? Deployment velocity. Not how fast you ship, but how fast the customer is actually live and seeing impact. Onboarding without friction, integration without a services engagement, no downtime, model flexibility, customization in days. The moat is no longer the product you built, but how quickly those products deliver results for the customer, and how quickly you can deliver more. So the conversation inside an AI startup should sound different than it did five years ago. Not "what else should we build," but what not to build, and where a missing product is an open invitation for the next startup to find its wedge. Focus hasn't died. It moved from the product to the customer. Calcalist כלכליסט
Experience
Honors & Awards
-
Included in the inaugural Midas List
Forbes
Languages
-
English
-
-
French
-
-
Spanish
-
Recommendations received
10 people have recommended Mike
Join now to viewOther similar profiles
Explore more posts
-
Laksheya Khanna
The easiest way to find VC roles before they’re even posted is.... Most people apply when a VC job goes public. By then, hundreds of applicants are already competing. The smarter way is to track fund lifecycles. VC firms usually hire when: 1️⃣ They are in the 3rd–4th year of their fund: That’s when portfolio companies mature and firms need more support. 2️⃣ They are raising or have just closed a new fund: More capital = more deals = more hiring. So instead of randomly applying, track which funds are deploying capital. That’s where opportunities appear before they’re announced. Ways to find this information: • Fund announcements and press releases • Platforms like Crunchbase, PitchBook, Tracxn • LP announcements and portfolio updates • LinkedIn activity of partners VC recruiting is less about job portals and more about understanding how funds operate. If you want a list of active investors and VC funds, comment “INVESTORS.” If you want advanced career playbooks, check this out: https://lnkd.in/gE9wy_sQ
195 Comments -
Simon Bøttkjær
When we started Evertrace, we had a hypothesis that software could outperform humans at sourcing investment opportunities. We felt that traditional tools fell short. They relied on founders showing up in the usual places, often far too late to create any real edge. By the time a company appeared, the opportunity was already crowded. So we started looking in more unusual places. We kept asking a simple question: what is the very first trace a founder leaves when they start working on something new? We built Evertrace from there. Today, more than 150 funds use Evertrace. Earlier this month, my co-founder Jacob Graubaek Houlberg went on etn. to talk more about our product, our vision, and how we see the early-stage venture game changing fast.
3 Comments
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More