Andreessen Horowitz

LLMs are quickly solving what we call "the messy inbox problem" —the overwhelming flood of unstructured data that slows down businesses across industries, from healthcare to freight. These AI models are automating top-of-funnel manual tasks, historically responsible for clogging operations, wasting countless hours, and creating bottlenecks in downstream processes. By synthesizing voice and text data from faxes, emails, PDFs, and call transcripts, and systematically extracting key information, LLMs are automating top-of-funnel manual workflows—emerging as a powerful wedge strategy for companies to streamline operations and gain a competitive edge. In his latest post, David Haber explores how AI applications, particularly those leveraging LLMs, are reshaping industries by owning the top of the data funnel, positioning startups to eventually displace legacy systems. Companies like Tennr, Heron Data, Happyrobot, and Eve are already capitalizing on this trend. Read more on how on how AI is transforming workflows and disrupting legacy systems: https://lnkd.in/gkAVzFwF

“This is a nightmare scenario.” Marc and Ben discuss the escalating U.S.-China competition in AI and robotics, highlighting concerns about China’s dominance in hardware—particularly Chinese drone manufacturer DJI’s share of the global drone market. For more, catch up on “The Ben and Marc Show” at 👉 https://lnkd.in/eEcuCgxE

Is Google’s search monopoly coming to an end? Alex Immerman shares how the end of Google’s search monopoly is near, as AI-native search engines gain traction with personalized, ad-free, conversational experiences. Hear more from Alex on this topic in the full episode on the a16z Podcast: https://lnkd.in/gYp3E7DV

The financial services industry—worth trillions—still has most U.S. dollars settling on “software built before the internet.” Jack Henry CTO Ben Metz is tackling one of fintech’s most entrenched problems: modernizing legacy banking infrastructure. Ben recently joined Angela Strange to break down: - Why outdated systems are slowing innovation. - How fintech builders can uncover opportunities by understanding core systems and their complexities. - What it takes to scale modernization—from engineering teams to infrastructure. Hear his advice for builders and more: https://lnkd.in/d9_8u7Ka

Accounting is running on “Reagan-era software”—but that’s about to change. While CPA numbers are declining, workload complexity and demand are only rising. AI has the opportunity to automate manual tasks, freeing accountants to focus on higher-value work. What’s the outcome? Faster service, lower costs, and a better experience for consumers. Seema Amble and Marc Andrusko break down why this is the perfect “why now” moment for AI in accounting. Link to the full video in comments.

What is debanking – and why is it happening? Simply put, debanking is when a law-abiding person or entity unexpectedly loses their banking relationship, potentially being kicked out of the banking system. Debanking (and “de-risking,” see more in the article) can occur without any apparent investigation, detailed explanation, or advance notice that would provide the entity enough time to move their funds. Most importantly: there is no due process, appeal process, or other recourse. Not every closing of a bank account is “debanking.” Banks can justifiably close their clients’ bank accounts for many reasons. They may also proactively choose to reduce their regulatory compliance costs and efforts by limiting their exposure to certain individuals, industries, or business models. However, this type of legitimate activity is not what’s raising red flags around debanking. While we can’t speak for an entire industry or particular interest, we as a VC in the crypto industry have seen, first-hand, at least 30 instances of debanking occur with our portfolio companies and founders over the last four years. Coinbase also publicly shared that they had uncovered at least “20 examples of the FDIC telling banks to ‘pause’ or ‘refrain from providing’ or ‘not proceed’ with offering crypto-banking services.” And recent FOIA results revealed details. There are likely many more cases. Read more about what debanking is, and why it matters, from a16z crypto: https://lnkd.in/g3N3h4Tu

Early-stage growth looks different for every company, but for Zapier, Gusto, and MongoDB, it started with staying close to their users: - Zapier focused on launching early, listening to customer feedback, and iterating quickly to refine their product. - Gusto spent hours with early customers, observing how they used the product to uncover ways to improve the experience. - MongoDB prioritized grassroots community-building through developer meetups, creating connections that turned attendees into early adopters. The co-founders share the lessons and strategies that shaped their early growth journeys on My First 16 with Seema Amble, a series featuring interviews with B2B founders about how they secured the first 16 customers that helped them go to market. Link to episodes in comments.

AI is redefining the potential of vertical SaaS. Markets once deemed too small, like laundromats, chiropractors, and veterinary clinics, are now becoming viable and scalable with the help of AI. As James da Costa and Angela Strange explain, AI can replace labor with software, boosting lifetime value per customer and reducing customer acquisition costs. This allows VSaaS companies to streamline entire workflows — from sales and marketing to back-office operations. Learn more about how AI is opening untapped markets in this week’s Enterprise newsletter 👇

There’s a giant opportunity right now, as AI turns capital into labor, says Alastair (Alex) Rampell. Capital no longer just buys engineers or hardware — it buys code that replaces or augments labor, unlocking new markets. On the a16z podcast, Alex, Angela Strange, and David Haber discuss digitizing systems of record, AI company “wedges,” and the near-term opportunities they see now.  Listen now wherever you get your podcasts, or at the link in our comments 👇

Thrilled to announce that we are leading the seed round for beeps, the free, on-call platform for Next.js developers. Beeps helps modern teams quickly identify issues, gain instant clarity on environment changes like deploys or external service problems, and keep users informed with ease. Looking forward to working with joey parsons and team.