ChatGPT Should Have Killed His AI SaaS. Instead, He's Nearing $10M

Summary of ChatGPT Should Have Killed His AI SaaS. Instead, He's Nearing $10M

by Omer Khan

53m•October 29, 2026
Listen to Episode

Overview of SaaS Podcast

In this episode of SaaS Podcast, Omer Khan speaks with the founder of Mosaic AI, an enterprise platform that automates technical support for B2B software companies. The conversation centers on how the company started as a broad enterprise AI assistant before ChatGPT commoditized that market, forcing a strategic pivot into a narrower but far more defensible wedge. The founder explains how he got to nearly $1M ARR, raised a $6.5M seed round, then rebuilt the company around a high-ROI support use case that is now pushing the business toward eight figures in ARR.

Business Overview

  • Product: AI platform for technical support automation
  • Customer: B2B software companies with complex products and large support teams
  • Scale:
    • Around 100 customers
    • Seven figures in ARR, nearing eight figures
    • Team of roughly 70–100
  • Go-to-market: Hybrid team split between Toronto and Tel Aviv
  • Pricing: Often six-figure deals, with some approaching seven figures
  • Value proposition: Reduce technical support costs by 20–30%+, with potential for 50–70% automation-driven savings in some accounts

The Founder’s Backstory

From BlueTail to Salesforce

  • The founder previously built BlueTail, a bootstrapped data aggregation company that functioned like an early version of ZoomInfo.
  • BlueTail was acquired by Salesforce, where he worked on Data.com as chief data scientist.
  • His co-founders went on to start other companies, while he later pursued a PhD in AI/NLP.

Academic Work That Shaped the Next Startup

  • During his PhD, he focused on reasoning for question answering.
  • He became involved in early transformer-era work, including research around BERT and the broader shift that led to modern LLMs.
  • That research convinced him that enterprise software would be transformed by AI, especially systems that could reason over company knowledge.

Why the Company Pivoted

The Original Vision: Enterprise AI Assistant

  • The company originally launched as Ask AI, an enterprise assistant connected to company data through roughly 100 connectors.
  • It worked well early on and sold to companies like Monday.com.
  • The team reached close to $1M ARR before ChatGPT changed the market.

The ChatGPT Shock

  • Once ChatGPT launched, the founder realized the broad enterprise assistant category was becoming a commodity.
  • Competing directly with heavily funded LLM giants like OpenAI and Anthropic/Claude would be unsustainable.
  • Rather than fighting in a huge red-ocean market, he looked for a “blue pond”:
    • a narrower market
    • with clear ROI
    • strong pain
    • and a path toward the larger enterprise vision

Why Technical Support

  • Technical support emerged as the right wedge because it is:
    • underserved
    • expensive
    • knowledge-heavy
    • deeply tied to internal systems and product understanding
  • The team reframed support as a cost center with measurable savings, which made the value proposition much easier to sell.

How Mosaic Sells

Founder-Led Sales in the Early Days

  • The first ~25 customers came through:
    • direct outreach
    • referrals
    • founder-led demos
    • SDR-as-a-service help
  • The first sales leader joined only in late 2023, after the company had already built meaningful traction.

Learning from Enterprise Conversations

  • The founder spent months talking to enterprises before raising capital.
  • Those conversations revealed the real pain:
    • fragmented knowledge across CRM, Confluence, Jira, Slack, tickets, and internal docs
    • support teams unable to answer complex customer issues quickly
  • One early design partner became a key proof point and helped validate willingness to pay.

The 5-Day Pilot

  • Mosaic uses a fast pilot to show value quickly:
    • the customer sends a large batch of tickets
    • Mosaic runs a diagnostic
    • identifies automations and knowledge gaps
    • deploys a working slice of the solution
  • The key metric is tickets per support rep, not just resolution time or ticket deflection.
  • In some cases, the company has seen as much as 80% growth in tickets handled per rep in a week.

Competitive Strategy

The Real Competition: Build vs. Buy

  • The founder says the biggest competitor is often the customer’s own internal dev team.
  • Because AI coding tools make building easier, enterprises are tempted to create their own support tooling.
  • Mosaic’s response is to sell a broader outcome:
    • not “we built a few features”
    • but “we can reduce support cost by 50–70%”

Why the Positioning Works

  • They’re not trying to win on feature parity alone.
  • Instead, they position themselves as the company that can:
    • run diagnostics
    • identify the automations that matter
    • connect to internal systems
    • and deploy the workflow end-to-end
  • This makes Mosaic harder to replace than a simple chatbot or internal prototype.

Best-Fit Buyers

  • The strongest buyers are often PE-backed companies, since they are highly motivated by:
    • cost reduction
    • efficiency
    • operational leverage

Key Takeaways

  • Start with the market, not the technology. Enterprise AI only works when tied to a specific, painful business problem.
  • Broad AI assistants can get commoditized fast. Narrowing to a use case can create defensibility.
  • Trust matters in fundraising. The founder told investors the truth when his co-founder left mid-round, and that honesty helped more than hiding it.
  • Enterprise buyers want outcomes, not demos. Talking about ROI and cost savings is more effective than showing off features.
  • Technical support is a strong AI wedge because it is expensive, complex, and full of data that AI can unify.
  • Speed matters. A short pilot can be enough if the product surfaces real, measurable value quickly.

Lightning Round Highlights

Best Business Advice

  • Automate everything — in the product, in operations, and in internal workflows.

Book Recommendation

  • Only the Paranoid Survive by Andy Grove

Best Money Spent

  • Building a strong, cohesive Toronto go-to-market team in an office/hybrid setup

Favorite Productivity Habit

  • Using AI tools as a personal assistant for:
    • travel
    • email
    • calendar
    • cost optimization

Outside-of-Work Passion

  • Jogging, usually 6–7 km most evenings
  • He likes doing it outdoors because it creates a sense of movement, challenge, and discovery rather than just “exercising in place”