The Future of Vertical AI

I had lunch with a founder last week who pitched me on their "AI for operations" platform. I stopped them 3 slides in. General-purpose AI isn’t cutting it anymore. DeepSeek’s January breakthrough told us something important: efficiency & performance can coexist a lot earlier than most people thought. Startups are now excelling not by scale but by focus: they’re building vertical AI that deeply understands the messy, high-stakes workflows in sectors like healthcare, finance, and defense. Specialization is the new competitive advantage. 3 patterns I’m tracking across successful vertical AI startups: First, they pick massive but high-friction and high-value workflows. “AI for sales” or “AI for operations” is too broad. What’s effective is focusing on urgent, complex processes, like: ConverzAI streamlining high-volume recruiting for staffing agencies Tennr automating messy admin work Second, they build more than model wrappers. They create proprietary feedback loops and data assets that compound over time. This instrumentation is what turns a one-off tool into a durable, defensible product. Third, they expand from beachheads of earned trust. They wedge into multi-billion-dollar industries by solving problems in the hardest, least glamorous corners. From there they earn the right to expand and unlock bigger TAM over time. Choose one gnarly high-value workflow and go deep. Otherwise you might get stopped three slides in too.

I have been an active investor for decades, and the growth in the market for vertical generative AI applications is beyond anything I have seen. Just in the accounting space, we went from tracking six companies that had raised close to $200M in funding to now tracking 56 that have raised $1.5 billion. In the legal world, over 100 early-stage generative AI startups have cropped up and raised $1.2 billion in funding to date. Industries like healthcare, finance, legal, and accounting are now booming with specialized GenAI companies, each focusing on specific functions and user needs. Across verticals, break-out companies are emerging with key traits: • 𝗥𝗣𝗔 2.0: Flexible and dynamic workflow automation for the enterprise, powered by LLMs • 𝗗𝗮𝘁𝗮 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴: GenAI tools for structured and unstructured data accelerate sector decision-making. • 𝗠𝘂𝗹𝘁𝗶𝗺𝗼𝗱𝗮𝗹 𝗟𝗟𝗠𝘀: Audio and text integration are improving interactions and documentation • 𝗔𝗜 𝗖𝗼-𝗣𝗶𝗹𝗼𝘁𝘀: Tools for junior professionals to increase speed, accuracy, and independence. Many of these trends come to life in our portfolio with companies like Collective, PilotDesk, Roe AI, Drillbit (YC S24) and e:cue. Our latest blog is a refresh of the market we first covered last year in ‘Generative AI in Vertical SaaS: Lessons from Reviewing 1,000 companies’ Special thanks to The LegalTech Fund, Andreessen Horowitz, and Menlo Ventures, whose valuable insights and research are referenced in the piece. It is an exciting time to track verticalized GenAI.

I’ve been wrestling with a fundamental question in AI: What happens to vertical AI solutions when models themselves become so good at reasoning and autonomous execution that they no longer need proprietary workflows? Right now, I completely understand the value of vertical AI—deep integrations, tailored workflows, domain-specific optimizations. These solutions are driving massive efficiency gains today, because models still need structured environments to deliver real business outcomes. But six months from now? Two years from now? The rate of improvement in general-purpose models is staggering. We’re not just talking about better responses; we’re seeing emergent capabilities that allow models to dynamically adapt, integrate on the fly, and construct workflows in real time without rigid pre-built architectures. If models can autonomously build and execute their own workflows, do we still need vertical AI solutions in the way we do today? Or do they get leapfrogged entirely? I’m not denying the value of vertical AI—it’s obvious right now. But I’m struggling to fully comprehend the long-term defensibility of proprietary workflows when the foundation models themselves are rapidly absorbing that capability. The integration layer today is where the magic happens. The question is, what happens when that magic becomes native to the model itself? Would love to hear perspectives on this. Are we underestimating the staying power of vertical AI? Or are we watching the early signals of its eventual obsolescence?

I used to be skeptical of vertical-focused startups—historically the TAM looked small next to broad horizontal plays. I’ve changed my mind. As others have written extensively, LLMs, AI Agents, and MPC lets founders push past traditional software boundaries while building deeper, defensible value. My partner Navin Chaddha wrote an excellent analysis on the case for vertical AI (link in comments). Here are the key takeaways: WHY VERTICAL AI WINS • Precision. In medicine, finance, or chips, a model that’s even slightly more accurate wins. • Fast ROI. Fine-tune on industry data with Agentic and MPC enabled workflows, deploy quickly, show savings or net-new revenue quickly. • Lower hallucinations. A narrow answer space built on clean domain data earns trust where “maybe” isn’t acceptable. • Built-in compliance. Treat regulation as a moat—design audit trails and privacy walls up-front. VERTICAL AI PLAYBOOK - Attack “hair-on-fire” use cases in a single domain; be the best, not the broadest. - Build proprietary data loops and generate synthetic data where necessary. - Start from the best open-source vertical model in your space and innovate on the last mile and integrations.  - Meet customers where they are—on-prem, VPC, or air-gapped; don’t let deployment friction slow adoption. AREAS OF INTEREST Healthcare care-coordination, underwriting and servicing, industrial maintenance, supply chain, audit, home services—all demand vertical precision today (and more). Each could be a billion-dollar company for a team pairing domain pain with AI advantage. BOTTOM LINE General LLMs are the platform. Vertical AI is where durable enterprise value accrues. We’re still in inning one—most sectors rely on manual workflows or legacy tools ripe for replacement. 

“Agentic” is going to be the word of next year, says Sarah Friar at OpenAI "In 2025, we will see the first very successful agents deployed that help people in their day to day" AI agents succeed not just because of great technology but because they align with users’ actual needs. Take private markets as an example: 👉🏻 Deal teams need IC memos. 👉🏻 IR needs quick DDQ turnaround. 👉🏻 Fund administration needs streamlined reporting. 👉🏻 Compliance teams need real-time monitoring. A generic AI agent won’t cut it here. It must: ✅ Integrate seamlessly with existing workflows. ✅ Deliver results in familiar, actionable formats. ✅ Work within the tools your team already uses. This is why many AI projects fail: too much time to deploy, excessive change management, and a lack of real alignment with business users’ expectations. What’s the solution? Specialized platforms, purpose-built for vertical tasks, that minimize friction and maximize impact. YC recently suggested that vertical AI agents could be the next SaaS revolution—and private markets are no exception. And that means a solution like Auquan: pulling domain-specific data from VDRs and relevant vendors, automatically generating credit overviews, fund reports, or ESG due diligence—all delivered in your team’s preferred format at the click of a button. The future isn’t just Agentic AI, its AI Agents tailored for your industry. #AI #Finance #RAG #PrivateEquity #PrivateCredit https://lnkd.in/eeQMUc8C

Are We Witnessing the End of #SaaS as We Know It? The rise of #Agentic Frameworks and Vertical #AI #Agents is shaking the foundations of traditional #SaaS solutions. For years, horizontal #SaaS #platforms have offered broad, one-size-fits-all tools, but they fail to deliver deep personalization and contextual understanding. Now, Vertical #AI #Agents are emerging—powered by Agentic RAG, domain-specific LLMs, and agentic workflows—to drive hyper-personalized product workflows for industry-specific use cases. What’s fueling this transformation? -The abstraction of AI through agentic systems. -Robust, general-purpose foundation models like GPT, Claude, PaLM, and LLaMA, now being augmented with domain-specific knowledge. These agents don’t just automate—they understand and adapt, delivering real-time insights and decisions tailored to industries like healthcare, finance, and supply chain. From predicting machine failures to optimizing patient care, hyper-personalization is becoming the new standard. This isn’t evolution—it’s disruption. Are we at the dawn of a new era where SaaS solutions fade and Vertical AI Agents take over? #AIAgents #VerticalAI #SaaS #AIInnovation #FutureOfSoftware #TechTrends #DigitalTransformation

scoot over SaaS agent-as-a-service is THE next frontier it will transform how businesses compete. I've spent years tracking tech adoption curves. what's happening with AI agents isn't just another trend, it's the beginning of a fundamental shift in how business services are delivered and consumed. for decades, enterprise companies had massive advantages: → custom software development  → specialized consulting teams  → dedicated operational staff  → sophisticated automation solutions small businesses couldn't compete. but AI agents are changing everything: → specialized business expertise, automated  → enterprise-grade capabilities at SMB prices  → 24/7 operation without staffing costs  → complex workflows handled autonomously this is no longer a theory, it's happening now. but first, let’s look back at the software world for a minute… horizontal platforms (like Microsoft) serve many needs across industries. vertical solutions (like Veeva Systems) go deep in specific sectors. AI agents are following the same pattern, but with a crucial difference: → vertical agents target enterprise needs with deep specialization  → horizontal agents address common SMB needs at scale  → the sweet spot? Marketplaces connecting them all my prediction? companies building AI agent marketplaces will be the next tech giants. why will marketplaces win? network effects → more agents attract more users → more users attract more agent developers → quality improves through competition cost efficiency → reduced marketing costs → shared infrastructure → streamlined discovery trust framework → verified performance → standardized interfaces → consistent experience look at enso, they're already connecting SMBs with specialized AI agents that would have required custom enterprise software just months ago. where do you fit in this new landscape? are you building solutions that will become agents? are you creating the marketplaces that will connect them? or are you preparing to transform your business with them? join Lighthouse where we look at possibilities brought by tomorrow's business tech. Link in comments 👇

Over the past few weeks, I’ve spoken with several entrepreneurs who are developing all-in-one, AI-powered vertical SaaS applications and making significant progress. In contrast, the previous generation of SaaS companies typically consisted of horizontal platforms that excelled in one market segment, such as marketing automation or sales engagement. These products were comprehensive, serving a wide range of industries. Over time, many of these applications moved upmarket, focusing on mid-market and enterprise clients. In addition to horizontal players, numerous vertical SaaS applications have emerged over the last two decades. These typically followed a playbook of targeting small to midsize businesses in specific segments before gradually moving upmarket. They focused on delivering the most valuable features for their target audience while avoiding overly broad functionality. With the rise of AI-powered software development, including low-code platforms and vibe coding, robust cloud computing resources, and mature open-source ecosystems, building large-scale software quickly has never been easier. As a result, entrepreneurs are now creating AI-powered vertical SaaS products that combine the functionality of multiple horizontal tools into a single, purpose-built solution for specific industries. Instead of small business owners needing separate tools for their website, social media, marketing automation, CRM, and sales engagement, a single system now provides everything they need, tailored to their vertical. These solutions are offered at a significantly lower price point with greater ease of use. Moreover, because these products are directly tied to revenue through lead generation, proposals, and new business, their ROI is clear. My recommendation to entrepreneurs is to identify a vertical they or a colleague know intimately and consider building a comprehensive application that replaces multiple existing tools for that target customer. By leveraging AI, cloud infrastructure, and open-source technologies, they can deliver a fully integrated solution at a fraction of the cost.

Over the holiday break, I had in-depth conversations with #VCs, #founders, and operators about how #AI will impact #SaaS. The prevailing sentiment is that AI will fundamentally disrupt SaaS as we know it. The argument is that AI will commoditize the creation of business workflow apps, thanks to its ability to generate code—and soon, workflows—with ease. A powerful AI agent could potentially “conjure up” any necessary workflow from scratch. Here are my key thoughts on this idea: 1) Timing of disruption remains uncertain: While #GenAI can generate code quickly, the broader engineering community is still skeptical about its ability to produce reliable, enterprise-grade code more efficiently than skilled engineers. Achieving stability and reliability in AI-generated workflows will require solving several open challenges in AI. 2) Intellectual property implications: If AI generates workflow code based on information gleaned from another #IP-protected SaaS product, would this constitute #IPInfringement? This remains an open question, much like debates around AI-generated content for copyrighted articles or images. 3) #VerticalAI agents will maintain strong moats: Just as vertical SaaS succeeded by leveraging domain expertise, domain-specific AI agents will continue to create value in a world of commoditized workflow apps. The more an AI agent understands specific business contexts, the more valuable it will become. 4) Systems of record are less susceptible to disruption: While AI may commoditize workflows and apps, enterprise systems of record (e.g., #CLM, #CRM, #ERP) will remain essential. These systems will serve as the foundational repositories for enterprise data that AI agents rely on to operate effectively. What are your thoughts on this? #AI #SaaS #MachineLearning #GenerativeAI #TechInnovation #AIInBusiness #EnterpriseTech #WorkflowAutomation #AIApplications #VerticalAI #BusinessTechnology #TechDisruption #IntellectualProperty #EnterpriseSoftware #BayAreaTech Sirion Eigen Technologies

In vertical SaaS, the very best companies landed with a product solving a specific, burning pain point that horizontal solutions solved badly and then became broad product suites over time (ServiceTitan, Veeva Systems, Doximity, etc). My partner Gordon Ritter calls this “layering the cake”. In vertical AI, the landing wedges are much more brittle. While the painpoints they solve are very real (and often unaddressable by SaaS alone), they’re also much more likely to commodotize in the near term as the tech advances. 👀 Therefore, the need to layer the cake in vertical AI is much more urgent than in vertical SaaS. 👀 The best vertical AI CEOs not only recognize the dynamic, but talk about it openly. Helps those around them keep up the speed of innovation and their eyes on the bigger prize.