We sit down with Kellan Carter, General Partner at FUSE, where he pulled back the curtain on what's really happening in AI investing, revealing why traditional SaaS metrics are misleading and how a new generation of founders is building completely different businesses.

The Revenue Quality Crisis Nobody Talks About

"For the last year, this came up in every single board meeting: how do I evolve my pricing?". Yes, founders chase vanity metrics, but the underlying unit economics are falling apart. Companies love their biggest customers but are losing money on them - a classic case of what Kellan calls "negative gross margin customers."

The wake-up call came when 20 AI voice companies launched in Q1, immediately cannibalizing each other's territory. "A minute is a minute is a minute," explains host Manny, "but customers aren't buying your drill - they're buying the hole it makes." This shift from selling features to selling outcomes is forcing a complete rethink of how software gets priced and sold.

The 2010s Playbook is Dead

The old SaaS formula was simple: raise capital, hire sales reps, hit quota, repeat. But when your product ties directly to customer outcomes, this linear scaling breaks down completely. "You almost have to erase everything you learned in the 2010s era," Kellan explains. The bridge between sales teams and product teams is collapsing as companies realize sustainable growth requires tighter coupling between what you build and what customers actually achieve.

Mission-Critical AI: The Ultimate Moat

While most AI companies compete on price and features, a small subset is building something entirely different: mission-critical infrastructure. Kellan highlights a 24-year-old founder building AI for 911 dispatching - technology that makes life-or-death decisions about emergency response priority.

"No one in public policy buys the cheapest option for 911.”. When your AI determines whether to dispatch SWAT or community police, quality trumps cost every time. These companies aren't just building software; they're becoming "the routing engine for how our world is governed from a safety standpoint."

The Great Professional Services Disruption

Imagine managing 2,000 operations people, except they're not people - they're AI agents working 24/7 without HR complaints or equity demands. Kellan sees freight forwarders, law firms, and insurance brokerages getting completely reinvented from the ground up. "What if a lawyer could handle 5,000 clients instead of 200?"

The question isn't whether this transformation will happen, but whether existing companies can adapt fast enough or if new entrants will capture the value. Companies with established relationships have trust and distribution, but startups have AI-first culture and 90% lower cost structures.

The BPO Acquisition Wave Coming

Traditional Business Process Outsourcing companies like Cognizant and Wipro aren't sitting idle as voice AI startups attack their core business. Kellan predicts a wave of strategic acquisitions: "It'll be really interesting to see who blinks first and buys a voice AI company, saying 'I know we built this for 100 years, but we're throwing it away and going AI-first.'"

The moment one major BPO makes this move, every competitor will follow within months, transforming an entire industry overnight.

Why VCs Now Ask "How Good is the Founder?" First

FUSE completely restructured their partner meetings to prioritize founder quality over growth metrics. "It's really easy to say I went from 0 to $1M last month, but how defensible is that?" The firm learned that early revenue can actually hurt long-term scaling when founders optimize for quick wins at the expense of sustainable architecture.

Quality founders think 10 moves ahead, building systems that can scale infinitely rather than hit short-term milestones. They'd rather grow slower and capture larger long-term outcomes.

The Vertical AI Investment Thesis

Kellan’s investment focus at FUSE centers on "vertical AI solutions" - companies where AI solves specific business outcomes rather than being the product itself. FUSE avoids the expensive model layer battle, instead targeting industries like logistics, insurance, and freight forwarding where AI can automate 70% of operational workflows.

The firm's dream investment? "The first software-enabled insurance brokerage. Software first, people second." Or a freight forwarder where AI agents handle all the document processing, carrier coordination, and customs management that traditionally required massive human operations teams.

What This Means for Founders

The companies winning this transition share common traits: they're tying pricing directly to customer outcomes, building defensible moats through mission-critical functionality, and optimizing for revenue quality over growth velocity. Most importantly, they're not trying to apply 2010s scaling tactics to 2020s outcome-based products.

As Kellan puts it: "Quality of revenue is evolving really, really fast in this AI world." The founders who understand this shift early will build the next generation of enterprise infrastructure. The ones who don't will become cautionary tales about the dangers of optimizing for the wrong metrics.

Companies mentioned

Aurelian - 911 dispatching AI voice agent
Owl - Insurance fraud detection and claims automation
Quandri - Insurance brokerage workflow automation
FreightMate - Voice agent for freight forwarding
Avante - HR/health tech AI company
Pictory - Video editing company
Carbon Robotics - Laser weeding systems for farmers
Symbiosis - Ad tech company acquired by DoorDash
Splunk - Data analytics platform
Boeing - Aerospace company
Blue Origin - Space company
SpaceX - Space exploration company
Amazon Kuiper - Satellite internet constellation
Starlink - Satellite internet service
Flexport - Freight forwarding platform
ServiceNow - Enterprise IT service management
Salesforce - CRM platform
Workday - HR and financial management software
Zuora - Subscription billing platform
Cognizant - IT services and consulting
Wipro - IT services company
Tata - Indian multinational conglomerate
Vapi - AI voice platform
Bland - AI voice platform
Nike - Athletic footwear and apparel
Costco - Wholesale retailer
Amazon - E-commerce and cloud computing
DataDog - Monitoring and analytics platform
DocuSign - Electronic signature platform

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

03: Simulation, Visualization, and Pressure with Andrea Austin, MD