💫 EntreConnect @ SaaStr Recap: How AI Agents Are Redefining SaaS—Insights from Founders & Leaders at Otter.ai, Aisera, EvenUp, Nebius & HeyBoss AI
EntreConnect Panel | SaaStr Week | May 2025
Last week, we hosted an electrifying EntreConnect session during SaaStr Week—bringing together unicorn founders, technical trailblazers, and infrastructure leaders to tackle one of the most urgent shifts in software: the rise of AI agents.
Over 1,000 founders, investors, and operators joined us to go beyond the hype, dig into real implementation, and debate what it truly takes to build enduring AI-native companies. From hilarious product war stories to deeply personal visions of avatar futures, it was one of our most unfiltered, high-signal events yet.
🎤 The Panel
Muddu Sudhakar, Founder & Founding CEO at Aisera – Serial enterprise AI founder, $180M+ raised, ex-ServiceNow, Splunk
Sam Liang, Founder & CEO at Otter.ai – $100M+ ARR, built AI meeting memory; ex-Google tech lead
Matt Chen, Head of ML of Document AI at EvenUp – Head of ML, building legaltech’s AI backbone; ex-AWS
Andrei Meganov, Head of GTM at Nebius – Runs GTM at NVIDIA/Accel-backed infra platform; ex-Yandex, BCG
Xiaoyin Qu, Founder & Chair at HeyBoss AI – Replaced herself with an AI CEO; ex-Run the World (acquired)
⚡ TL;DR – Key Takeaways
🪦 SaaS isn’t dead—but it’s evolving fast with agents augmenting or replacing traditional workflows.
🧮 Outcome-based pricing is winning as token-counting and seat models fall flat, especially with SMBs.
🧠 Moats are shifting—speed, trust, and user lock-in are more critical than ever.
🤖 Personal AI avatars could be the next consumer platform, blending memory, personality, and utility.
💀 SaaS Isn’t Dead—But It’s Evolving Fast
Is SaaS dead? Muddu Sudhakar didn’t mince words: “SaaS is done. Don’t even come pitch a SaaS tool unless it has agents.” To him, the age of agent-powered systems has already arrived—and traditional SaaS is getting left behind.
Sam Liang offered a more tempered take: “SaaS will be augmented by agents—not replaced. You can still build great SaaS businesses, but the next layer is outcome-driven AI agents on top.”
Matt Chen noted a fundamental shift in how humans interact with software: “AI is now your co-worker. It’s not just integrating into your workflow—it becomes your workflow.”
Andrei Meganov framed the evolution as a move toward bionic organizations: “SaaS used to be about shipping tools users could figure out. Now, we’re building hybrids—AI agents embedded in workflows, working alongside humans. But customers still don’t have the skills to build these systems themselves. So SaaS will evolve into more of an integration and orchestration layer.”
Xiaoyin Qu echoed that sentiment from the SMB side, adding: “Our users don’t want to learn prompts or workflows. They want results—fast, frictionless, and done for them. That’s where agents win.”
The verdict? SaaS isn’t dead—but it's morphing fast into a new agent-native stack.
💸 From Seats to Outcomes: Rethinking Pricing
Pricing models are undergoing a seismic shift—and every speaker agreed that the seat-based model is losing relevance.
Sam explained how Otter grew by offering incredible value at little cost upfront: “We gave away 600 minutes for free. Once you build the habit, you’re hooked.” Over time, the product's usage-based value—not seats—drove monetization.
Xiaoyin was even more direct: “Token-based pricing is nonsense to SMBs. They just want one prompt to get everything done. We make money when their website makes money—that’s it.”
Muddu hammered it home: “No one’s funding seat-based SaaS anymore. VCs want outcome-based models. Period.”
From an infrastructure angle, Andrei explained how usage-based pricing is also evolving. “Inference is becoming the new metric. With agents, it’s not just about how much compute you rent—it’s about how fast and reliably that compute powers decisions. Our pricing is now tied directly to delivery quality and responsiveness.”
The message? If your GTM assumes old school SaaS logic, it’s time to rewrite the playbook.
🏰 Building Moats in the AI Agent Era
“What’s your moat?” used to be a question about IP or data. Now? It’s about habit, trust, and speed.
Sam Liang shared how Otter built a powerful moat through “the meeting knowledge base.” With eight years of company conversations stored and searchable, switching away becomes nearly impossible. “It’s like email—you don’t just casually move your history.”
But Muddu pushed back on the idea of any lasting moats: “Post-ChatGPT, two months is a long-term roadmap. You win by moving fast, iterating fast, and copying fast.”
Xiaoyin added a fresh take: emotional loyalty. “Your AI agent can become your trusted friend. If it deeply understands your business, knows your voice, delivers outcomes—why would you leave it?”
Moats now come from relationships, reflexes, and retention—not patents.
📈 Go-to-Market: Scaling from $0 to $100M+ ARR
How do you get to $100M ARR in this new world? Otter’s Sam Liang shared their viral playbook: start free, make it sticky, and go bottom-up. “We gave away 600 free minutes. Once you have 20 meetings saved, it’s hard to switch.”
Xiaoyin, in contrast, launched HeyBoss with a bit of flair. “A competitor tried to steal our traffic by registering a fake domain. I called them out publicly—and it went viral. We gained users fast.”
Muddu stressed one thing above all: “Don’t waste time on product feedback from people who won’t pay. Focus on serious buyers—customers who have budget, urgency, and willingness to commit. Get revenue in the door early. Iterate later.”
Andrei added the enterprise lens: “We started Nebius by going straight to large-scale AI teams—offering them GPUs when no one else could. But now, we’re building community, hackathons, and long-term engineering goodwill.”
🔐 Earning Trust in Regulated Industries
Matt Chen highlighted how trust—not just security—is key in high-stakes industries like law. “We’ve invested in tooling to track data lineage, enable auditability, and empower our customers to trust the system.”
He also noted a trend toward edge AI: “Keeping data close to the customer builds confidence. It’s not just about compliance—it’s about control.”
As expected, Muddu didn’t hold back: “If you’re not making money, no one cares about your security posture. Focus on ROI. Governance comes last.”
🏗️ Infrastructure: Built for Agents, Not Just Models
Andrei brought a crisp framework: “Training equals investment; inference equals revenue.” AI agents are supercharging inference demand—and reshaping where compute must live.
“People hate waiting on agents,” he said. “Latency matters. We’re bringing data centers closer to users, not just to reduce cost, but to keep agents responsive and usable.”
Muddu punctuated this with a callout: “There’s no killer agent yet. Whoever builds it will be the iPhone moment for this era.”
🧑🎤 AI Avatars: Memory, Personality & Retention
If SaaS had users, agents may have relationships. And AI avatars might be the ultimate personalization.
Xiaoyin described Astra, HeyBoss’s avatar CEO: “She’s not just a helper—she’s a brand. Customers build loyalty to her.”
Sam Liang revealed Otter is building an avatar of himself. “It’s trained on eight years of my meetings and writing. It already gave a press interview at Bloomberg.”
He added, “One day, I hope my kids can still talk to my avatar when I’m gone.” The room went silent for a beat.
Andrei countered with a philosophical caution: “If you offload your least favorite tasks to an agent, how are you different from Severance? I’d rather eliminate them altogether.”
The future of avatars is equal parts practical, emotional, and ethical—and it’s coming fast.
🔍 Finding Conviction from 0 to 1
The final question of the night struck a personal chord: how did these founders find conviction before PMF?
Sam said simply: “I built Otter because I couldn’t remember my meetings. I needed it. That was enough.”
Xiaoyin offered perspective: “Many of the ‘overnight successes’ launched four times. They don’t tell you that part.”
Andrei recommended pre-selling: “Find one real customer, even if small, and build side by side. Brutal feedback beats blind belief.”
Muddu, of course, kept it blunt: “Get on a plane. Go to St. Louis. Sell. You don’t need advisors—you need buyers.”
🧭 Final Takeaway: Welcome to the Agent-Native Era
The era of SaaS as we know it is giving way to something more dynamic, embedded, and outcome-driven. AI agents are not just tools—they’re co-workers, sales reps, designers, and even CEOs. The companies that win in this new landscape won’t be those that build yet another product, but those that embed agents into the fabric of work, relationships, and decision-making.
Whether it was Sam’s vision of every user having a personalized memory avatar, Xiaoyin’s AI-led company that literally fired her, or Muddu’s bold calls for speed and outcomes over process and moats—one thing was clear:
Founders must move faster, think deeper, and design for trust, personalization, and performance.
We’re no longer iterating on software. We’re co-creating with software.
🙏 Thank You + What’s Next
A massive thank you to our powerhouse panelists for bringing unfiltered insights, real conviction, and hard-earned wisdom to the EntreConnect stage. 💥
To everyone who joined us—thank you for your energy and curiosity. This was one of our richest discussions yet, and we’re just getting started.
👉 Follow us on LinkedIn and Luma for future event invites and exclusive takeaways
🗓️ Join us at our next gathering—we’d love to see you there.
Stay tuned for our upcoming LinkedIn posts spotlighting photos 📸 of the incredible audience who joined us at the event!
Let’s keep building, exploring—and creating the next wave of AI-native greatness. 🚀
💬 LinkedIn Challenge: Share, Learn, Connect
Thank you to everyone who participated in our LinkedIn Challenge! We're thrilled to feature the most engaging and inspiring post (link here), giving our community a chance to celebrate and learn from the experience. We also truly appreciate everyone who shared their best moments and insights with us!
🗓️ Upcoming Events
Follow us on Linkedin and Luma to stay updated on future events!
June 13, San Francisco, CA | How Agents Will Revolutionize the Future of Data @AWS GenAI Loft, Join us for an exclusive evening with leading AI and data visionaries—including the Founder & CEO of Neon ($130M raised, unicorn acquired by Databricks for ~ $1 billion), Gen’s Chief AI Officer, Founder & CEO of Datastrato, and Aisera’s VP of AI.
June 26, Mountain View, CA | AI & Agents: Most Exciting Use Cases & Opportunity for Enterprise Tech, join us for a provocative, content-rich presentation and Q&A with Muddu Sudhakar, unicorn founder and founding CEO of Aisera & Investor
July 30, San Francisco, CA | Infrastructure & Agentic Workflows with C-Suite from MongoDB, Databricks and more.
🧳 Career Opportunities
We are looking for volunteers! Join us to help shape the future of our founder community, drive meaningful impact, and grow alongside some of the most inspiring entrepreneurs out there:
👉 Check out this LinkedIn post and this one for more opportunities from our amazing panelists' companies!
 | A guest post by
|