In recent months, enabled by recent advancements in AI, vertical SaaS startups have started using industry data to build out a powerful set of specialised workflows. Historically, venture dollars have flown disproportionately toward horizontal SaaS companies addressing incredibly large markets (I.e. CRMs for SMBs, etc.). Recently, however, a number of early-stage VCs have realised that some of the vertical SaaS companies are building specialised software that is so powerful that it will have very high penetration levels.
Let's look at one of many such examples. VIC.AI has built customised industry-specific models on the basis of proprietary accounting data sets. With large proprietary data sets, VIC is able to deliver a set of incredibly powerful services:
- Ingest invoices in any format to decode and extract key information
- Automate invoice processing based on a certain set of best practices based per invoice type
- Route payment processing into different flows based on VIC's intelligence (customised for invoice type)
By using proprietary data, VIC is able to bring automation to accounts payable while saving time and reducing the risk of fraud. This is a classic example of using industry-specific data sets to build high-precision and high-efficiency workflows that can achieve very high levels of penetration within a specialised workflow.
Amidst the noise in the broader generative AI space, which currently makes up less than 1% of total cloud spend, I believe vertical SaaS startups will form one of the most interesting layers of early winners.