Insights Unlocked: “AI Employees and the Future of Work” with Surojit Chatterjee

Podcast: Insights Unlocked by UserTesting
Date: November 3, 2025
Host: Mike Mace (UserTesting)
Guest: Surojit Chatterjee (Founder & CEO, Emma; former CPO at Coinbase, VP/Product at Google)
Duration: ~47 minutes

Episode Overview

This episode explores the transformative role of AI employees (digital AI-driven agents) in shaping the future of work. Surojit Chatterjee shares insights from his experience building large-scale tech products (at Google and Coinbase) and now as the founder of Emma, a company developing universal AI employees. The conversation traces the profound shifts AI is bringing to the workplace—from redefining mundane tasks and company structures, to championing customer empathy, rapid experimentation, and reskilling. The tone is candidly optimistic, focusing on actionable strategies for leaders navigating this paradigm shift.

Key Discussion Points & Insights

1. Surojit Chatterjee’s Journey and the Genesis of Emma

[02:12 – 04:56]

• Surojit’s background: Computer science, product leadership at Google (Mobile Ads, Shopping), Coinbase.
• Two core insights from his career:
  • Rapid technological advancement—AI was already integral at Google and Coinbase, now even more so with generative AI.
  • Enduring drudgery for employees—Despite progress, most employees spend 50%+ of time on “mundane, soul-crushing” work.
• The founding hypothesis of Emma:
  • Future teams will comprise both human and AI employees.
  • “What if we could build a universal AI employee that could take any form...and take a lot of that mundane work, soul-crushing work away?” – Chatterjee [04:34]

2. AI: Colleague or Competitor?

[04:56 – 08:37]

• Historical perspective (computers, internet, industrial revolution)—each wave sparked fears of job loss, but ultimately created new categories of work.
• Chatterjee predicts “leaner” companies, but more companies overall and new opportunities to solve humanity’s toughest challenges.
  • “...yes, there will be leaner companies...but there will be more companies because there are more interesting things to do.” – Chatterjee [07:14]
• Optimistic outlook: Technological advances consistently raised overall prosperity instead of eliminating human purpose.

3. The Big Shift: From Tools to Thought Partners

[08:37 – 12:58]

• Earlier tech (e.g., Mac) was seen as a tool. Generative AI is more like a colleague, assistant, or “thought partner”.
• AI now can:
  • Conduct deep research, answer complex questions, offer advice, and bounce ideas.
  • Example: AI providing second opinions on medical lab results for family—boosts confidence and comfort but doesn’t replace experts.
    • “…It can [now] be a thought partner. You can bounce ideas off against each other…” – Chatterjee [09:32]

4. Deep Customer Empathy & Rapid Experimentation

[12:58 – 16:09]

• Traditional enterprise software forced companies to adapt to rigid software processes.
• With AI, software can now adapt to the customer/company’s workflow, even generating dynamic UI and automations on the fly.
  • "We believe we are writing software that adapts to how the enterprise works, not the other way around." – Chatterjee [13:46]
• Payment and engagement models changing: Usage-based pricing and adaptable software are replacing seat-based/licensing models.

5. Implications for Legacy Software Companies

[16:09 – 19:39]

• Large, established tech companies risk obsolescence if they fail to restructure products for the AI paradigm (adapt software to users, not vice versa).
• Historical analogy: Many 90s software titans vanished in the cloud era.
• "Every company that started before 2020 is a legacy now. Unfortunately, the timeframe has squeezed…" – Chatterjee [17:33]
• Warning to incumbents: Refactor and adapt or risk “being stale.”

6. The Agentic Business Transformation (ABT)

[19:39 – 22:48]

• Each company must examine every function (sales, HR, support, marketing, finance) to decide:
  • What is best for humans versus AI?
  • How does mano-a-mano and hybrid (human + AI) collaboration work?
• Vision: Human-AI collaboration becomes “normalized”, much like self-driving cars are now routine.
  • "You'll be working with a bunch of AI colleagues. You may have an AI manager; you may be managing AI agents or AI employees." – Chatterjee [21:38]

7. How to Get Started: Practical Steps for Companies

[22:48 – 26:36]

• Start early and learn—don’t wait for the perfect use case.
• Educate and reskill employees for prompts, delegation, and collaboration with AI.
• Manage both over-expectation (“AI will replace everything tomorrow”) and fear (“AI will never work / it will break everything”). Reality is in between.
  • "Everybody needs to be reskilled... it's as interesting or as complex as working with another human or managing employees." – Chatterjee [23:54]
• AI needs human-level feedback, nuanced delegation, and framing—much like developing a new manager skillset.

8. Early Use Cases with Real Traction

[27:18 – 29:49]

• Emma aims to automate any workflow (“horizontal platform”). Early adopters most keen in:
  • Customer support: Many companies starting here.
  • HR: Complete internal employee experience—Emma replaced HR in Emma’s own company.
    • “Internally at Emma we have no HR. We only use Emma.” – Chatterjee [28:14]
  • Finance: Data processing, reconciliation, procure-to-pay, order-to-cash.
  • Sales & Marketing: Writing proposals, slide decks, RFPs, contract analysis.
• Prediction: In 5-10 years, every function will be touched by AI employees.

9. Debunking the Productivity Paradox

[29:49 – 36:30]

• Debate: Does AI actually boost company productivity? Academic skepticism vs. case-study optimism.
• Reason for confusion: Many companies simply “inject” AI into old processes without fundamentally changing workflows.
  • Example: A large company kept 9 separate expense mailboxes (with human and AI watching each); recommended “design one seamless process using AI” instead.
• Key conclusion: Big productivity gains require “large-scale re-architecture of the organization” rather than piecemeal automation.
  • "If you transform the process itself, you will see 50%, 60%, 80% improvement." – Chatterjee [32:41]
• Industry will see “confusion and learning” before “step-change” leapfrogs for AI-first organizations.

10. The Importance of Transparency & Trust

[36:30 – 39:12]

• Trust in AI, like trust in new employees, is earned through consistency and transparency:
  • Show decision rationale (“Why was the AI’s decision made?”), adapt to user feedback, and behave consistently.
  • "Same principles apply with AI... So AI has to be transparent in terms of explainability." – Chatterjee [37:35]
• Feedback loops and transparency pivotal for reliable collaboration.

11. Actionable Takeaways for Leaders

[39:12 – 45:40]

• Non-tech or legacy companies: Begin with easy wins (customer support, HR), but don’t stop there. Work iteratively, and plan for deep transformation.
• Don’t treat AI as a blunt cost-cutting tool; use it to “make your employee investments more productive, more valuable.” [41:05]
• Reskilling and employee buy-in are essential: Top performers at Emma are the heaviest AI users.
  • "Guess what? Our highest performing employees also use AI the highest.” – Chatterjee [43:19]
• Cultural humility: Recognize AI is a partner, not a rival. “We have a partner now.”

Notable Quotes & Memorable Moments

• On paradigm shifts:
  “Every company that started before 2020 is a legacy now. Unfortunately, the timeframe has squeezed...”
  — Surojit Chatterjee [17:33]
• On hybrid work:
  "You'll be working with a bunch of AI colleagues. You may have an AI manager, you may be managing AI employees. So it will be a hybrid workplace."
  — Surojit Chatterjee [21:38]
• On adoption and upskilling:
  “It’s almost like you have an infinite supply of new employees without hiring a single one. Because your current employees have the most context... If you made them more efficient, you could grow your business much faster.”
  — Surojit Chatterjee [41:50]
• On experimentation and employee usage:
  “Our most highest performing employees also use AI the highest.”
  — Surojit Chatterjee [43:19]
• On adapting old mindsets:
  “The biggest change we need to make is in ourselves and our own expectations.”
  — Mike Mace [44:41]

Timestamps for Key Segments

• [02:12] Surojit on his tech journey and why he started Emma
• [04:56] Examining fears around AI and job impact
• [08:37] Generative AI as a colleague vs. traditional tools
• [12:58] Building customer-adaptive, malleable software with AI
• [16:09] Implications for legacy software companies
• [19:39] Agentic business transformation: redefining company structure
• [22:48] Practical steps for initiating AI transformation
• [27:18] Early AI use cases with high adoption
• [29:49] Why productivity gains lag: process vs. tech changes
• [36:30] Transparency and building trust in AI
• [39:12] Action items for companies & employee upskilling
• [43:19] Data: AI use vs. high performance at Emma

Episode Summary and Tone

With clarity and optimism, this episode provides both a roadmap and a rallying call for leaders facing the AI transformation era. Surojit Chatterjee’s perspective is unflinchingly optimistic but rooted in a clear-eyed assessment of change: AI will neither eliminate human value nor leave old company structures untouched—it will, instead, drive an inventive hybrid era of work. Productivity, transparency, and adaptability are the watchwords, and the leaders who start early, experiment broadly, and put employee development first will be those who thrive.

For further details and resources:
Visit emma.co or usertesting.com/podcast