1Mby1M Virtual Accelerator AI Investor Forum: With Venktesh Shukla, Founder and Managing Partner at Monta Vista Capital (Part 3)

Sramana Mitra: Are there any other portfolio companies that you would like to discuss? What you’ve just provided is a good framework – Generative AI hallucination is acceptable without human in the loop if it’s non mission critical. If it is mission critical, we need human in the loop. But even with human in the loop, human in the loop is necessary, but not sufficient to control hallucination, especially in medical situations, where it really can be dangerous.

So those are good heuristics. Is there another case study that we can do?

Venktesh Shukla: After we made that investment and after ChatGPT came, you must have seen at least a 100 companies claiming to be using generative AI. Last year, we did not invest in a single Generative AI company because most of the ones that we saw were really thin wrappers on top of ChatGPT. We didn’t think that there was a fundamental, sustainable differentiation over a long period of time.

So quite interestingly enough, last year we did not invest in a single generative AI company. We invested in AI companies, but not in generative AI companies. Now, we are getting ready to actually invest in two companies. One is in the medical space and one is in the sales marketing space.

Sramana Mitra: And you’re not yet interested in discussing those two that you’re about to invest in?

Venktesh Shukla: No, we could talk about it.

Sramana Mitra: I would like to know about what passed the due diligence.

Venktesh Shukla: One company is essentially trying to come up with the next generation of electronic health records (EHR). Today, all large practices use either Epic or Cerner for EHR. If you talk to any doctor, the number one frustration that a doctor or a nurse has is the number of clicks they have to make and the amount of time they have to spend entering data.

Sramana Mitra: Yes, old architecture, clunky user interface.

Venktesh Shukla: This is where generative AI comes in. When dealing with a patient, you have the entire history and test results. For instance, say the patient is an 18-year-old male with specific medical conditions. Instead of presenting the doctor with twenty checkboxes, generative AI allows the system to narrow down options to just two or three highly relevant choices.

This way, the doctor can select one, guiding them down a path without needing 14 clicks just to reach the prescription stage. If an 18-year-old patient has a certain disease history, and test results confirm this, the AI contextually presents only the relevant options. The doctor selects a prescription or treatment, streamlining the entire process efficiently.

Sramana Mitra: Have you been able to move Epic out of medical systems?

Venktesh Shukla: You know, I think Epic is the 800-pound gorilla, and large systems don’t make changes as easily.

Sramana Mitra: Huge exit barrier.

Venktesh Shukla: They’ll need to follow a similar path to what Salesforce did when it started. Salesforce initially targeted small and medium-sized businesses. When I was at Magma, we were one of its first customers, and at that time, Magma was a company with around $20 million in revenue. As Salesforce accumulated more of these smaller customers, it had to continuously add features to meet their needs. Over time, it built out enough functionality to transition to enterprise customers, because it had developed all the features that anyone could ever demand.

Sramana Mitra: OK, and what is the other one?

Venktesh Shukla: We have not finished with the diligence for the other one. It’s in the sales and marketing space, where it enhances the efficacy of BDRs. BDRs typically write tons of emails every day to customers, and they try to customize it as much as they can. This company takes as much information as it can from public records and the company records, to essentially suggest a draft.

Sramana Mitra: So it is a business development rep (BDR) writing assistant.

Venktesh Shukla: Right. It doesn’t need to have a human in the loop, because nothing is catastrophic if you’re sending a hundred of those emails. Even if two to five of those are wrong, it doesn’t matter.

This segment is part 3 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Venktesh Shukla, Founder and Managing Partner at Monta Vista Capital
1 2 3 4