We are in the midst of one of the largest platform shifts in history. Just as the personal computer slashed the cost of compute and the internet cut the cost of distribution, generative AI is rapidly reducing the cost of productivity.
We believe that, in short order, the tasks below (and more) will be done entirely by AI. 
Today, every enterprise is grappling with how to “add” AI to their business, while entrepreneurs who are building with AI are unlocking efficiencies never before possible. Whether you’re an executive aiming to ramp up productivity in your company through AI or an entrepreneur looking to start a new business, there is opportunity wherever you see a preponderance of these tasks.
While AI companies have made significant progress — especially in infrastructure layers, such as LLMs, vector databases, and software development tooling — most of the real, application-level breakthroughs have yet to be built. In the near future, companies will fully productize AI, making the technology increasingly effective for a wide range of use cases.
How to leverage AI right now
How should you think about the immediate opportunity for AI, whether you’re incorporating the tech into your company or building new AI-based solutions? One potential framework: Tasks contained to a specific role? AI agent. Tasks contained to an industry function? AI vertical software. Tasks spread throughout your workflows? AI horizontal software.
- AI agents replace basic roles: Highly repetitive, deterministic tasks or workflows. This includes basic data collection, simple customer service tasks, and document synthesis. These are highly automatable and can be replaced by genAI agents, enabling companies to hire fewer employees or move to higher order tasks.
- AI vertical software drives efficiency in teams: Connecting workflows, which requires decision making with some ambiguity. This includes data entry into company customer relationship management (CRM) systems, processes that require interaction across domains, or workflows that incorporate multiple software tools. GenAI-optimized workflows or industry-specific software solutions will empower employees to spend time on higher order tasks, rather than rote work.
- AI horizontal software elevates knowledge workers: Admin and support tasks. These include generating meeting notes, drafting follow-up emails and reminders, reviewing code/wireframe, and more. These tasks weigh on even the highest level executives. Here, we need horizontal genAI tools that free up workers to focus on creative thinking and complex decision making.
Let’s look at specific examples in each category and the potential broader implications:
Agents replace basic roles
Jobs that predominantly consist of the tasks above are likely to be almost entirely subsumed by AI agents. For example:
- Compliance analysts: Know Your Business (KYB) processes are often still conducted manually. To confirm that a business applying for a bank account is real, humans might search its address on Google Maps to check that the Street View image matches the description. Today, much of this work can be done via AI tools like Parcha or Greenlite. Similarly, agents from Norm.ai can check that marketing copy is compliant with regulations like the SEC marketing rule, replacing human review.
- Sales development reps (SDRs): Sales teams spend much of their time searching for and qualifying leads, inputting data, drafting outreach, and managing pipeline reports — rather than actually selling. New players like 11x are automating these tasks, so sales teams can spend more time on high value work and face time with customers.
The implications:
- Companies may hire fewer people in some roles (Klarna AI customer service reps, for example, are already doing the work of 700 full time agents), or potentially more people in others (with greater top-of-funnel, for example, a company could support more account executives).
- Workforces will become much more scalable. Today, many industries struggle to respond to seasonal or market dynamics, such as hiring more customer service reps during peak travel season or more mortgage loan officers when rates drop. AI agents are easy to “hire.”
- The incumbent software that currently supports every back office function — CFO, CRO, CCO — will eventually add AI. We’re already seeing this with products like Salesforce’s Einstein platform. In time, however, new companies that start out as agents targeting a specific pain point are likely to expand. By tackling additional tasks and workflows, such as sales outreach, sales enablement, or lightweight customer relationship management (CRM), these new entrants may eventually compete with incumbents to become the preferred system of record.
Vertical solutions drive efficiency in moderately skilled teams
In addition, the software supporting *industry specific* functions will dramatically transform with AI. If your (legal, finance, medical practice etc.) teams are still performing any of the above tasks around your software, consider new solutions. A few examples:
- Tennr automates away the manual processes of patient care coordination for healthcare practices. The AI makes sense of documents (which are often faxed and handwritten by doctors) and extracts the relevant patient information into the practice’s Electronic Health Record system (EHR). This allows the practice to deliver the right care more promptly.
- A Loan Origination System (LOS) for mortgages, like Vesta, will automatically ingest documents, accurately extract relevant data, and communicate with borrowers to ask for any missing paperwork. Casca, a LOS for commercial loans, already purports to increase productivity by 10X and slash manual work by 90%.
The implications:
- New AI companies will emerge to automate tricky pain points or detail-oriented processes where previous software attempts were too error-ridden or costly. In areas where you may have had to hire “business operations,” there may now be an opportunity for an AI solution.
- Systems that were once considered too dangerous to “rip and replace” are now vulnerable. Prior substitutes were 2x better than incumbents; the new AI-first competition is 10x better, thus increasing buyer receptivity. Incumbents that require trained consultants for lengthy implementations, where customers must file tickets for feature fixes and wait weeks for updates, are particularly vulnerable (including ERPs, CRMs, banking and insurance core systems, and more).
Horizontal solutions elevate knowledge workers
Are you at least 20% more productive due to AI? You probably should be. We all encounter tasks in our day-to-day work that should be automated by software — first drafts, summarization, note taking, searching. However, often these tasks are spread throughout the day or across multiple activities, making it seem less worthwhile to put in the effort to automate them. Now AI provides such gains that AI-first, horizontal tools are emerging to cover the gaps. For example:
- Granola.so will listen in on your meetings and summarize notes. Eventually, AI could draft your follow-up emails and perform other action items.
- Zapier’s AI allows you to create automated, bespoke workflows across apps. For example, you could create a tool that would automatically take contacts from your calendar and populate your personal CRM, including up-to-date information about each contact.
The implications:
- Even in these early innings, knowledge workers are reporting productivity gains greater than 20%. Whenever you think, maybe software should be doing that for me? The answer is a definitive yes.
How much more productive could you and your organization be?
While we are still in the early days, AI can already provide dramatic productivity gains. If you have special domain expertise about a process, workflow, or role that can be made more efficient with AI, we’d love to talk to you. If you’re a larger company looking to get started — or just move faster — taking a hard look at where these automatable tasks exist in your organization is a good start.
