goodfin (YC W22)

.@DoorDash is acquiring @Instacart. Two of the biggest names in delivery are joining forces. Instacart built its business around grocery technology, retailer relationships, and advertising. DoorDash built a massive consumer delivery network. Now, they’re bringing those businesses together. 𝗧𝗵𝗲 𝗯𝗿𝗲𝗮𝗸𝗱𝗼𝘄𝗻 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀 Instacart announced a definitive agreement to be acquired by DoorDash. The deal brings together DoorDash’s logistics network and Instacart’s grocery technology and retail partnerships. The combined business aims to expand grocery access, improve retailer tools, and create more earning opportunities for shoppers. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: Instacart’s value extends beyond grocery delivery. Its advertising business, retailer integrations, and grocery data give DoorDash capabilities that would take years to build organically. For DoorDash, the acquisition could mean more frequent orders, deeper retailer relationships, and additional advertising revenue. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: For private-market investors, this is another reminder that exit opportunities extend beyond IPOs. Strategic acquisitions can offer a path to liquidity when an established platform sees more value in buying infrastructure than building it. The deal also raises a bigger question: how much consolidation is still ahead in delivery and commerce?

Former OpenAI executive Kevin Weil just raised $215M to bring AI and robotics into drug discovery. His new startup, @warpsciences , has emerged from stealth at an $865M valuation, with plans to build automated chemistry labs in the U.S. Here's the breakdown from the news:   • $215M Series A led by Lightspeed Venture Partners.   • $865M post-money valuation at launch.   • Investors: IVP, Mayfield, Lux Capital, Premji Invest, and Menlo Ventures.   • Founding team: Former leaders from OpenAI, Meta, the White House, and Terremoto Biosciences.   • The technology: AI models and robots designed to synthesize and test drug compounds.   • Initial focus: Lead optimization, a critical stage in identifying promising drug candidates. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 Drug discovery still relies on expensive, time-consuming physical experiments. Warp aims to automate laboratory workflows and run experiments around the clock. Its U.S.-based labs could reduce dependence on overseas chemistry contractors. Every experiment generates data that can improve subsequent AI-driven research. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Warp’s $215M raise reflects a growing investment thesis: AI’s next major opportunity may be in automating physical work, not just digital tasks. But building robotic labs is capital-intensive. Warp will need to prove that automation can deliver faster experiments, competitive costs, and reliable scientific results. The opportunity is significant. So is the execution risk. What are your thoughts? Source: Warp Sciences, Lightspeed Venture Partners, Bloomberg, TechCrunch

.@AnthropicAI just filed for an IPO. Its numbers are staggering. The Claude maker is heading to Nasdaq under the ticker ANT, potentially setting up one of the largest IPOs in history. 𝗧𝗵𝗲 𝗻𝘂𝗺𝗯𝗲𝗿𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀:   • $16.3B in H1 2026 revenue, up from $1.2B a year earlier.   • $693M operating loss, down from $2.1B.   • $80.7B net loss, largely driven by noncash accounting adjustments.   • 11,900+ customers spending at least $100K annually.   • $1.5T valuation reportedly being discussed by investors, down from earlier $2T expectations.   • 50.1% of voting power will remain with Anthropic’s seven founders. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: Anthropic’s revenue has grown more than 13x year over year. But rapid growth comes with enormous infrastructure costs, complex financing, and questions about long-term profitability. For the first time, public investors will get to price those trade-offs in a frontier AI company. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: Anthropic could establish the valuation benchmark for the entire frontier AI sector, including OpenAI. The real question isn’t how much revenue AI can generate anymore. It’s how much of that revenue becomes sustainable profit.

.@WeAreLegora raises $550M at an $8.8B valuation, just seven months after its last $550M round. The legal AI startup has increased its valuation by nearly 57% since March 2026, backed by explosive revenue growth. 𝗧𝗵𝗲 𝗯𝗿𝗲𝗮𝗸𝗱𝗼𝘄𝗻 𝗰𝗼𝘃𝗲𝗿𝗲𝗱 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀: Legora secured a $550M Series E led by Atreides Management. Its valuation jumped from $5.6B to $8.8B in seven months. Annual recurring revenue (ARR) surpassed $200M, up from $1M in less than two years. The platform now serves 140,000+ legal professionals across 2,400+ organizations. New investors include Fidelity, J.P. Morgan, Baron Capital, and Capital Group. 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 Legora added its second $100M in ARR in under six months. More than half of the AmLaw 50 now use Legora. In-house legal teams accounted for over 40% of new customers in Q3. The funding will accelerate U.S. expansion and development of its agentic legal operating system. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: 𝘄𝗵𝗮𝘁 𝘄𝗲'𝗿𝗲 𝗸𝗲𝗲𝗽𝗶𝗻𝗴 𝗮𝗻 𝗲𝘆𝗲 𝗼𝗻𝗲 Legora's growth shows what investors increasingly want from enterprise AI: measurable revenue, adoption by major institutions, and expansion into new customer segments. With $200M+ in ARR, Legora is proving that specialized AI platforms can scale well beyond the initial hype.

.@salesforce is acquiring @Braze for $5.66 billion, paying a 58.1% premium to expand its AI-powered marketing capabilities. The all-cash deal strengthens Salesforce’s position in real-time customer engagement. Here's the breakdown:      • Deal value: $5.66B in enterprise value.   • Offer price: $47.50 per share, all cash.   • Premium: 58.1% above Braze’s previous closing price.   • Customer base: 2,770 customers, including Canva, Venmo, and Instacart.   • Expected closing: FY2028, subject to approvals.   • Financial outlook: Salesforce’s FY2027 guidance remains unchanged. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: 𝘄𝗵𝘆 𝗯𝘂𝗶𝗹𝗱 𝗶𝘁 𝘄𝗵𝗲𝗻 𝘆𝗼𝘂 𝗰𝗮𝗻 𝗮𝗰𝗾𝘂𝗶𝗿𝗲 𝗶𝘁 This brings Salesforce a stronger marketing cloud, adding Braze’s real-time personalization and cross-channel messaging to its marketing stack. Combining Salesforce’s customer data with Braze’s AI-driven engagement tools strengthens Salesforce’s position in enterprise marketing software, supporting its AI expansion strategy through established technology rather than building from scratch. Talk about a competitive edge. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: 𝘄𝗵𝗮𝘁 𝘄𝗲'𝗿𝗲 𝘄𝗮𝘁𝗰𝗵𝗶𝗻𝗴 Salesforce is paying a 58.1% premium for Braze’s technology and customer base. The real test is whether integrating Braze into Salesforce’s ecosystem can generate enough growth to justify the $5.66B deal.

.@Tenstorrent raises $1.5B at a $6.2B valuation as investors bet on alternatives to Nvidia. The Toronto-based AI chip startup is building a vertically integrated computing platform spanning chips, servers, and software. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀? Tenstorrent reportedly raised $1.5B at a $6.2B valuation. The company develops AI processors using the open-source RISC-V architecture. Its technology combines custom silicon, high-performance computing systems, and software designed to reduce dependence on Nvidia’s ecosystem. Tenstorrent is led by CEO @jimkxa, a veteran chip architect who has worked at Apple, AMD, Tesla, and Intel. 𝗪𝗵𝘆 𝗱𝗼𝗲𝘀 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿, 𝘆𝗼𝘂 𝗮𝘀𝗸? AI infrastructure is diversifying. Nvidia dominates the market, but customers and investors are looking for alternatives. Software is the real battleground. Competing on chip performance alone isn’t enough. Developers need accessible tools and reliable software support. Open architectures create opportunities. RISC-V offers greater flexibility for companies developing customized computing systems. Capital requirements remain enormous. Building competitive AI hardware demands substantial investment in engineering, manufacturing partnerships, and software development. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: Tenstorrent’s funding highlights investor appetite for alternatives to Nvidia. But the bigger test isn’t raising capital or building faster chips. It’s winning customers, securing production capacity, and building a software ecosystem developers actually want to use.

.@greptile raises $50M at a $750M valuation. It now reviews more than 1% of GitHub pull requests daily. The YC-backed startup is building an independent AI-powered code review and testing platform, addressing a growing bottleneck in software development. I𝗳 𝘆𝗼𝘂 𝗺𝗶𝘀𝘀𝗲𝗱 𝗶𝘁, 𝗵𝗲𝗿𝗲'𝘀 𝘄𝗵𝗮𝘁'𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀.   • Greptile secured $50M in Series B funding, led by Menlo Ventures.   • Investors include Jump Trading, Wing, Operator Collective, Benchmark, Y Combinator, Browder Capital, and Initialized Capital.   • The company previously raised a $25M Series A in September 2025.   • Greptile says it reviews and tests more than 1% of all pull requests opened on GitHub daily.   • Thousands of engineering teams use its platform, including Nvidia, Coinbase, Decagon, and Vanta.   • Menlo Ventures investor Deedy Das is joining Greptile's board. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: AI coding tools like Cursor and Claude Code are accelerating software development. But producing more code also creates more work for engineers responsible for reviewing and testing it. Greptile is targeting that bottleneck. Its platform independently validates code changes, helping teams identify bugs before they reach production. As companies deploy multiple AI coding agents, an independent review layer could become a critical part of their development infrastructure. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁? AI is making code generation faster. Code validation is becoming the next investment opportunity. Greptile's $750M valuation reflects investor interest in the infrastructure needed to make AI-generated software reliable. Its reported GitHub penetration demonstrates significant adoption. The next test is whether Greptile can consistently reduce review time, catch critical bugs, and help engineering teams ship faster.

.@modal's valuation jumped from $4.65B to $15.75B in just five months. The AI infrastructure startup has raised $750M in Series D funding, led by Accel, after crossing $1B in annualized revenue run rate. 𝗪𝗵𝗮𝘁'𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀   • Modal raised $750M at a $15.75B post-money valuation, more than triple its May 2026 valuation.   • Accel led the round, with participation from Kleiner Perkins, Menlo Ventures, General Catalyst, Redpoint, Lux Capital, and Bain Capital Ventures, among others.   • The company reported 15X+ year-over-year growth, with platform usage tripling since its Series C five months ago.   • Modal's Sandboxes product, which enables developers and AI agents to execute code, has reached a $300M annualized run rate.   • The platform has launched nearly 3 billion sandboxes and can support up to 1 million running concurrently. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 AI infrastructure is attracting substantial capital. Modal's valuation has increased by $11.1B since May, backed by rapidly growing demand for compute. AI agents are becoming a major revenue driver. Sandboxes now account for approximately 30% of Modal's reported annualized run rate. Developers need more than GPUs. Modal provides infrastructure for model training, inference, and executing code without requiring companies to manage the underlying compute. Revenue growth is supporting higher valuations. At $15.75B, Modal is valued at roughly 15.75X its reported annualized run rate, although that figure is not the same as full-year recognized revenue. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Modal's $300M Sandboxes run rate is the number to watch. As AI agents move from answering questions to executing code and completing tasks, the infrastructure needed to run them is becoming a significant business. Modal is already converting that demand into revenue.

Goodfin is a proud partner of the AI LAX Summit during LA @TECHWEEK_ by a16z, bringing together the people building, investing in, shaping, and deploying the next generation of AI. The Summit convenes AI leaders, builders, founders, investors, creative pioneers, policymakers, and technical leaders to explore how humans and AI agents will work together — and what the agentic era means for technology, business, creativity, infrastructure, and society. Across keynotes, panels, fireside conversations, workshops, demos, and discussions, the Summit will explore: * The Culture Machine — AI for Everyday Life * Agents at Work — Enterprise Modernization & Yield * Governance, Trust & Security — Fiduciary Oversight in the Agentic Era * Frontier Infrastructure — Power, Compute & Deep Tech * The LA Innovation Agentic Economy - New Economic Paradigms & Venture Capital The day will include speakers and leaders from organizations including @Google , @GoogleDeepMind, @fivetran , @circle , Goodfin, @Hasbro / Sixthwall, and others. Join us hearing perspectives spanning AI, technology, entertainment, finance, infrastructure, aerospace, policy, and venture capital — alongside opportunities to connect with the builders and leaders shaping what comes next. 10:00 AM - 6:00 PM Friday, October 16, 2026 Event details: ailaxsummit.com

Uber's former CEO just secured $1.7B for his robotics venture. Uber is among the investors. Travis Kalanick's @Atoms raised $1.7 billion in equity funding led by Andreessen Horowitz (a16z), bringing together investors including Uber, Bain Capital, and Fifth Wall. 𝗪𝗵𝗮𝘁'𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀   • a16z led the round, with co-founder Ben Horowitz joining Atoms' board.   • Uber participated in the funding, reportedly investing $100 million.   • Atoms is consolidating businesses including CloudKitchens and autonomous mining technology company Pronto.   • The company is developing industrial automation across three sectors: food production, mining, and transportation. 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀? Atoms is targeting specialized robots built for specific industrial tasks rather than general-purpose humanoids. Its existing businesses give it access to commercial kitchens and mining operations where automation can be deployed. Robotics requires significant upfront investment in hardware, deployment, and maintenance. The $1.7B round provides capital to expand those operations. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Atoms already has operating businesses and industrial infrastructure. The question now is whether it can turn automation into a profitable, repeatable business across multiple sectors. For investors, deployment costs, customer adoption, and operating margins will matter more than the size of the funding round.

An AI model that doesn’t generate text just raised $870 million. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀? @typesafeai has raised an $870M Series A at a $7.5B valuation, led by Andreessen Horowitz (a16z). Sequoia Capital and DCVC participated in the round. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗱𝗼𝗲𝘀: TypeSafe’s model, Jev, makes structured, probability-based decisions instead of generating text. Jev is designed for tasks like classification, routing, and software automation, where speed and cost matter more than conversation. The company claims roughly one-third of Fortune 500 companies already use Jev. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁: Most enterprise software doesn’t need an AI chatbot. It needs fast, reliable decisions at scale. TypeSafe is betting that specialized models can deliver those decisions more cheaply than general-purpose AI. At a $7.5B valuation, converting enterprise usage into recurring revenue will be the real test.

.@TeamOratomic just hit a $5.4B valuation, three months after its Series A. The quantum computing startup raised $475M in Series B funding, bringing its total funding to $775M. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀? Valuation: Jumped from $1.5B to a reported $5.4B. Investors: ARCH Venture Partners, Khosla Ventures, Spark Capital, Index Ventures, General Catalyst, and Bezos Expeditions. Technology: Oratomic is developing a fault-tolerant quantum computer using neutral atoms. Breakthrough: Its research suggests useful quantum computing could require just 10,000–20,000 qubits, compared with earlier estimates of around 1 million. Timeline: CEO Dolev Bluvstein is targeting a working system by 2030. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Oratomic’s valuation has more than tripled in three months, despite having no commercial quantum computer yet. Investors are backing its approach to error correction, one of quantum computing’s biggest technical hurdles. The next test is whether Oratomic can turn its research into a reliable, commercially viable machine. $775M in funding gives it substantial resources, but the engineering challenge remains.

.@tryramp's valuation just jumped $16 billion in four months. The fintech startup has reportedly raised $1.85 billion at a $60 billion valuation, up from $44 billion in June. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀? Funding: $1.85B, led by Dragoneer and Thrive Capital, with participation from Founders Fund. Valuation: $60B pre-money, up 36% since June. Customers: 70,000 businesses as of June, up from 50,000 at the start of 2026. Revenue: More than $1.5B in annualized revenue as of June. Founded in 2019, Ramp started with corporate cards and expense management. It now offers payments, procurement, accounting automation, and AI-powered finance tools. For context, its former rival Brex was acquired by Capital One for $5.15 billion earlier this year. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 At $60 billion, Ramp is valued at roughly 40 times its June annualized revenue run rate. Investors are paying for more than a corporate card business. They’re betting Ramp can become the primary platform businesses use to manage spending, payments, and financial operations. The next test is whether Ramp can grow revenue fast enough to justify that premium.

AI is leaving the lab and moving into systems that fly, sense, manufacture, and operate. Getting there takes founders, operators, government, regulators, and capital working from the same map. That’s what ecosystem builders like @AILA_Community make possible. They create the connective tissue: the introductions, shared context, and trust. We think about access the same way. The most important frontier companies will spend most of their growth years private. Relationships that matter form long before a round is ever announced. Supporting the communities where those relationships begin isn’t adjacent to our work. It is our work. We’re glad to support AI LA’s Mission Critical at THE BR-DGE x DTI during LA @Techweek_ by @a16z: an evening with leadership from the U.S. Space Force, senior military leaders, regulators, investors, and operators from aerospace and defense startups and unicorns. On the table: • autonomy and mission operations • the path from prototype to deployment • AI-assisted engineering and advanced manufacturing • government adoption and procurement • the role of venture in what comes next. If you’re building where software meets the physical world, this is the room. We look forward to building alongside you. 📅 Wednesday, October 14 ⏰ 3-7pm 📍 THE BR-DGE | El Segundo 🎟️ Register: lnkd.in/g6ZrC2Zz #LATechWeek #AILA #DefenseTech #Aerospace #DeepTech

.@ereborbank went from $1.1B to $7B in deposits in six months. Now it’s reportedly raising $1.5B. Founded by Palmer Luckey and Joe Lonsdale, Erebor is building a bank for startups, venture-backed companies, and the crypto industry. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄𝘀   • $7B+ in deposits, up from $4.06B in June and $1.1B in March.   • 500+ new customers reportedly added in Q3.   • $1.5B fundraising target at an $8B pre-money valuation.   • Stablecoin conversions: Erebor has reportedly restricted free USDC and USDT redemptions after trading firms exploited its zero-fee policy. 𝗪𝗵𝗮𝘁'𝘀 𝘁𝗵𝗲 𝗯𝗶𝗴 𝗱𝗲𝗮𝗹: 𝗧𝗵𝗲 𝘀𝘁𝗮𝗯𝗹𝗲𝗰𝗼𝗶𝗻 𝗹𝗼𝗼𝗽𝗵𝗼𝗹𝗲 Erebor initially allowed customers to convert USDT and USDC into dollars at a 1:1 rate without fees. Trading firms including Wintermute and Galaxy Digital reportedly took advantage. They could buy USDT below $1, redeem it at Erebor for a full dollar, and pocket the difference. Erebor has since introduced fees and limits for smaller customers while retaining free conversions for qualifying large depositors. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Erebor’s deposit growth is impressive. But deposits are liabilities, not revenue. The bank must maintain sufficient capital against its growing balance sheet. Its reported 12% leverage requirement makes additional fundraising particularly important. The next question is profitability. Can Erebor turn $7B in deposits into a sustainable banking business without relying on expensive incentives?

CRACKED happened. Here’s what stuck. 🌶️ During SF @Techweek_, we gathered a room of founders at Marlowe for a stacked breakfast and honest, closed-door conversation with @bonatsos (Verdict Capital) and @pramodsharma (Napkin AI), moderated by @VirajAla21, sharing the real insights on what building really looks like. A few things worth carrying into your Monday - for founders and non-founders alike: 𝗦𝘁𝗿𝘂𝗴𝗴𝗹𝗲 𝗶𝘀𝗻’𝘁 𝘀𝘂𝗳𝗳𝗲𝗿𝗶𝗻𝗴. Struggle is hard work where you’re still learning and going deeper. Suffering is effort with no new insight. If you’re suffering, change direction. 𝗚𝗮𝗺𝗶𝗳𝘆 𝘁𝗵𝗲 𝗻𝗼. You can’t prep for a 90%-rejection fundraising rate until you live it. The mental shift: go out and collect nos until they stop stinging. 𝗧𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗶𝘀 𝗮 𝗽𝗿𝗼𝗱𝘂𝗰𝘁 𝗼𝗳 𝘁𝗵𝗲 𝗽𝗿𝗼𝗱𝘂𝗰𝘁. Napkin spent two years experimenting without pull. Things changed when the product clicked and creators started sharing it organically — not when they bought more ads. (Fun fact: napkin.com wanted $500,000. Pramod bought napkin.ai for $20. Unintended ROI.) 𝗣𝗲𝗿𝗳𝗲𝗰𝘁 𝗶𝘀 𝘁𝗵𝗲 𝗲𝗻𝗲𝗺𝘆 𝗼𝗳 𝘁𝗵𝗲 𝗴𝗼𝗼𝗱. We've heard this time and time again, but a product is never going to be completely perfect, and you must get comfortable with imperfection. Benchmark your product against existing market options rather than an idealized standard to know when it is ready to sell. 𝗗𝗼 𝘁𝗵𝗲 𝗯𝗼𝗿𝗶𝗻𝗴 𝗽𝗮𝗽𝗲𝗿𝘄𝗼𝗿𝗸 𝗲𝗮𝗿𝗹𝘆 — 𝗶𝘁’𝘀 𝘄𝗼𝗿𝘁𝗵 𝗺𝗼𝗿𝗲 𝘁𝗵𝗮𝗻 𝘆𝗼𝘂 𝘁𝗵𝗶𝗻𝗸. The most expensive founder mistakes are quiet ones: the 83(b) you forgot to file, the QSBS clock you didn’t know had started. Most founders find out the rules after the window to act has closed. The ones who win on this started paying attention at incorporation, not at acquisition. That last one is close to our heart — QSBS can shield a meaningful share of your gains from federal tax when you exit — but only if your equity was structured and held the right way, early. It’s exactly why we offered every founder in the room a complimentary QSBS diagnostic from Goodfin: a fast read on whether your equity could qualify, and roughly what it could be worth to you and your team. And equally important, everyone walked away with our CRACKED hot sauce. Hand picked by Viraj for its spice profile: for bold founders and builders only. The caliber of the room was the point. Thank you to everyone who came hungry and curious — and to Viraj for making it real. More is coming. If this is your kind of room, stay close. #sftechweek #founders #cracked

A great resource for private markets investors navigating embodied AI.

.@anduriltech is investing $3.7 billion to build a new defense manufacturing facility in Maryland. The defense technology company plans to develop Arsenal-2, a 2 million-square-foot facility at Sparrows Point. 𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗡𝗲𝘄𝘀   • Anduril has announced a $3.7B private investment in Arsenal-2, aimed at expanding U.S. defense manufacturing capacity.   • The facility is expected to create over 3,100 permanent jobs and support more than 14,000 jobs directly and indirectly.   • The project could receive public incentives covering up to 13% of the investment.   • Anduril projects more than $2B in annual economic output once the facility is operational. 𝗪𝗵𝘆 𝗜𝘁 𝗠𝗮𝘁𝘁𝗲𝗿𝘀 Anduril has built its business around autonomous defense systems, software, and advanced military hardware. Arsenal-2 would give the company additional manufacturing capacity to meet growing demand from U.S. defense customers. The investment also highlights the capital requirements of defense technology. Developing advanced systems is one part of the business. Manufacturing them at scale requires factories, equipment, skilled workers, and long-term contracts. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 Defense tech startups are increasingly competing on their ability to manufacture, not just innovate. Anduril’s $3.7B commitment shows how much capital companies may need to move from developing defense technology to producing it at scale. For private market investors, manufacturing capacity, government contracts, and execution timelines are becoming important factors in evaluating defense tech companies.

Tab emerged from stealth at a reported $300M valuation after raising a pre-seed round. The San Francisco startup is building a personal AI agent that operates through iMessage and WhatsApp. 𝗖𝗼𝗺𝗽𝗮𝗻𝘆 𝗗𝗲𝘁𝗮𝗶𝗹𝘀 Brennan Erbz, Stafford Schlitt, and Ammar Amdani founded Tab. SV Angel, Valar Ventures, and American Spirit backed its pre-seed round, according to Traded. Unlike most AI assistants, Tab does not require users to open a separate app. Users send it a message through iMessage or WhatsApp. The agent can then carry out tasks such as booking flights, calling businesses, drafting emails and researching purchases. Tab has not disclosed the size of the pre-seed round. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 A $300M valuation at pre-seed sets a high bar for a company just emerging from stealth. Tab’s approach also tests a simple distribution idea: AI agents may not need their own apps. iMessage and WhatsApp already have billions of users and established daily habits. If Tab can reliably complete tasks from a text message, it can reach users without asking them to adopt another interface. For consumer AI agents, distribution is as important as the underlying model.

AI agents can make purchases. Most fraud systems still assume there’s a human behind the transaction. .@beltichq has emerged from stealth with a $7.3M seed round led by Norwest, bringing its total funding to roughly $8.8M. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗱𝗼𝗲𝘀 Beltic is building security infrastructure for businesses dealing with AI agents that act on behalf of customers. Their system is called “Know Your Agent” (KYA). It helps businesses identify an AI agent, verify who is behind it, and determine what it has been authorized to do. An agent may have permission to compare flights, but not to purchase one. Traditional fraud tools rely on signals like device fingerprints and behavioral patterns. Those checks become less reliable when software is navigating and transacting on a customer’s behalf. 𝗚𝗼𝗼𝗱𝗳𝗶𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁 AI agents taking actions on the internet add a new identity and permissions problem. Businesses already verify people through KYC and companies through KYB. If autonomous agents begin handling more transactions, they may need their own verification layer too. Beltic is betting KYA becomes part of that infrastructure.