Suggestions
Andrew Steele
Partner at Activant
Andrew Steele is a highly skilled professional with a background in economics and psychology, having pursued a Bachelor of Arts degree from Harvard University.
He has held key positions at prestigious organizations, including serving as a Partner and part of the Investment Team at Activant Capital.
Andrew also co-founded IvyMinded, showcasing his entrepreneurial spirit and leadership capabilities.
Highlights
Voice can feel like an old medium in B2B software. Why bother with calls when you can message? Why bother with call centers when you can have structured data input into forms?
Well, we see voice as a powerful application of AI, a means to reduce friction for the customer, expand the surface area of opportunities for a business, and ultimately play an important role in establishing the new guard systems of record.
A few examples we're seeing in the real world today:
Stuut (accounts receivable automation.) Tarek and the team have launched AI calling functionality where Stuut's agent payers, is able to securely pull all relevant information from disparate systems, and can execute on bill payments. Stuut is already seeing strong outcomes and major lifts in NPS for their customers (after all, it can be a little awkward speaking to another human about a bill).
Avoca AI (vertical AI for HVAC, electrical, & plumbing). Apurva and Tyson are rapidly scaling AI agents to drive new revenue for their customers. One job for an HVAC business can be $30k+ in value and job schedules can be tight - every missed opportunity is very costly. Avoca's platform is closes leads across all channels, connects systems across the value chain, but also coaches reps (think vertical Gong) to train existing sales personnel.
Sana - enterprise knowledge management. Joel and the Sana team are using call and meeting recordings to supplement data in an enterprise's knowledge base. By taking note of all the unstructured information shared in internal meetings, rich data can be automatically added.
We see some common themes in the best AI platforms:
🔹They're not just customer support / call centers, they're drivers of net new revenue or value (in addition to customer support).
🔹 They don't just offer voice, but offer context. They connecting disparate systems and data on the backend and add previously unstructured data from calls.
🔹 They earn trust either by announcing to the customer that they are an AI agent, or by elegantly announcing they're in the meeting. No customer wants to be babied on AI, and trust is everything in external and internal relationships.
🔹 They elevate humans to provide the elite customer experience by focusing them on the highest value areas, providing context, and training reps. Humans are great at establishing relationships with other humans, and if we play this out the baseline will rise to feel everyone has high status.
🔹 They earn (or are in the process of earning) not just read but write access to existing systems (e.g. pre-filling the CRM), and more importantly create net new data to earn the right to be a new system of record.
We're spending a lot time on the emergence of new systems of record and as a follow up to our initial research piece on conversational AI: https://lnkd.in/ePQHJ-8t
If you're applying voice AI to your platform or have seen cool examples, please reach out, we'd love to chat
Rough thought: a common AI topic for founders and investors these days is defensibility. Different layers of the stack have different considerations, but the more we see in the application layer, the more clear it is that defensibility will come in the usual ways: becoming a system of record. Why, because AI in its current form can create moat breaking moments of magic and earn the right to take on incumbents head on. A few thoughts:
🔹Most software (or at least the best software) is about convincing users to input data into some form of a database - whether it's consumers posting on social or businesses filling in fields in a a CRM.
🔹We all know that data is gold (and we're long gold), but the obvious challenge for challengers is it's hard to move from incumbents.
🔹We all know the feeling - why are we still using the ugliest, clunkiest software for many core business workflows.
🔹We saw the first wave with multi-tenant SaaS, but there’s still a lot of mediocre software owning and bottlenecking critical workflows.
🔹So a pretty front end, while incredibly important, isn't enough to convince key stakeholders to put in the work to move data from one database to another, and go through the pain of convincing other people to do the same.
🔹To actually get a business buyer to switch requires going through a LOT of red tape.
🔹To break this inertia you need moments of magic, or as Alex Rinke from our portfolio company Celonis puts it - "lightning in a bottle".
🔹Turns out AI is really good at creating moments of magic.
🔹When you turn up and a series of fields have been prefilled, and a real workflow has been completed automatically, the percieved effort required to get value from any platform immediately disappears.
🔹Instead as a user you can focus on the value potential, the incredible UX, the time you now have to do things of value rather than be a data jockey.
🔹The result - we can now take on incumbents in markets that previously felt impenetrable.
🔹We have a window of time to use AI to wow customers, build disciples, and build new data moats.
🔹And we can combine that with unstructured data at a rate never seen before.
🔸We've been seeing some incredible examples in the real world, and have several unannounced investments that we’re excited to share soon.
I'm going to share a short essay on this theme soon with some case studies. If you're seeing the same trend (or even better, building it), if you're also excited, or if you're just downright skeptical, please shoot me a message. Would love to collaborate!
#ai #software
