About a month ago, Andrew Pignanelli’s LinkedIn post launching Co-Founder, an AI chief-of-staff that runs your business with you, went viral.

In this episode, we go behind that moment to unpack the tech, product, and philosophy behind it. Andrew shares how The General Intelligence Company (GIC) is building agent systems that remember context, make long-term decisions, and act autonomously.

We dig into the company’s two-agent architecture, memory layers, and the culture of experimentation that shaped the build. Plus, what it means to “run vision-first” in a space moving this fast.

--

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

--

Key Takeaways

• Context is everything. AI without memory or goals can’t make reliable decisions.

• Experimentation is a superpower. Great teams ask for outcomes (“make it 10% faster”) and let engineers prototype freely.

• Context engineering is the new frontier. The challenge isn’t retrieval — it’s choosing which 20 paragraphs of context matter.

• Knowledge graphs over vector search. GIC builds structured relationships between people, projects, and data.

• Taste matters as much as tech. Human judgment now shapes what feels “right” in an agent’s behavior or interface.

• Vision-first beats incremental. GIC bets on big missions that attract deep believers and top talent.

--

⏱ What’s Covered

(01:15) Andrew’s journey from Velvet to The General Intelligence Company

(02:27) The idea behind Co-Founder

(03:55) The viral launch and early traction

(05:30) Why “memory” is the last step to general intelligence

(12:04) How GIC structures agents

(15:30) Knowledge graphs and grounding memory in real data

(19:18) Letting engineers experiment

(23:55) Context engineering and “taste” as product principles

(28:43) The building of Co-Founder

(35:10) Why GIC runs vision-first

(41:27) How Co-Founder makes entrepreneurship accessible

(48:42) AI, jobs, and ownership

(54:56) What’s next for Co-Founder and the path to general intelligence

-- 

Resources & References

• LangChain — Framework for building agentic and retrieval workflows

• Devin — AI software engineer for dev teams

• Claude (Anthropic) — Model family for long-context reasoning

• GIC Blog — Deep dives on memory, context, and architecture

--

Where to find Andrew Pignanelli:

• LinkedIn: andrewpignanelli

• X/Twitter: @ndrewpignanelli

• Website: The General Intelligence Company

👀 They’re hiring: generalintelligencecompany.com/careers

Where to find Haroon Choudery: 

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/ 

• X: https://x.com/haroonchoudery

Christine Y. Cruzvergara is Chief Education Strategy Officer at Handshake, where she partners with colleges and universities to help millions of students find their first jobs and internships.

Before joining Handshake, Christine served in senior roles at Wellesley, Georgetown, George Mason, and George Washington University, building scalable career education programs that bridge campus to career.

--

What You’ll Learn

1. The real skills employers want from graduates in the AI era
2. How Handshake AI and the new Fellowship program train students to work with AI, not against it
3. Why your first job should be treated as a learning role instead of the final destination
4. How to network effectively when AI now screens the first round of resumes
5. Why domain expertise (even outside CS) is becoming more valuable in AI-driven roles
6. The mindset shift from “find a job” to “build experience”
7. How higher ed and employers are collaborating to create new pathways for students
8. Christine’s advice for graduating students entering an uncertain job market

--

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

--

What’s Covered

- (00:00) Christine’s background and Handshake’s mission 

- (03:28) Why Handshake launched Handshake AI and how it works 

- (06:00) The Handshake AI Fellowship: who it’s for and what it teaches 

(09:38) What skills and majors are in demand for AI-related roles 

(13:14) The current job market for graduating seniors 

(15:17) How to set expectations and approach your first role strategically 

(16:36) The new importance of networking in an AI-screened hiring world 

(20:00) How Handshake is using AI internally to boost productivity and insights 

(39:48) The Future AI Workforce Alliance and how higher ed is evolving 

(54:14) How to apply and resources for students

--

References and Resources

Yoodli AI: https://yoodli.ai

Handshake’s 2024 Graduate Job Market Report: https://joinhandshake.com/docs/network-trends/class-of-2024-graduation.pdf

--

Where to Find Christine Y. Cruzvergara

• LinkedIn: https://www.linkedin.com/in/christinecruzvergara/
• Handshake: https://joinhandshake.com
• Handshake AI Fellowship: https://joinhandshake.com/fellowship-program

Where to Find Haroon Choudery

• LinkedIn: https://www.linkedin.com/in/haroonchoudery
• X: https://x.com/haroonchoudery

Most analytics tools tell you what happened. James Evans argues that in the age of AI, tools need to tell you what to do next.

In this episode, James shares how Amplitude is rethinking analytics with AI -- from anomaly detection that surfaces fixes, to session replay that could replace brittle event taxonomies. We dig into the cultural and technical shifts teams need to make to become AI-ready: embedding copilots into real workflows, running hack sprints instead of hackathons, and proving adoption through prototypes, not PRDs.

—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—

What You’ll Learn

1. Why peer demos beat training, and how healthy FOMO drives faster AI adoption.
2. Why prototypes are the new PRDs.
3. How hack sprints outpace hackathons.
4. Why AI fluency must start at the top (execs, we're looking at you).
5. Why distribution matters more than features.
6. How to manage the consistency risk — with governance and guardrails when everyone codes.

—

What’s Covered

(00:00) James' introduction and the Command Bar acquisition

(09:27) How to build background agents that find signals and trigger fixes

(10:21) Prototype a single onboarding agent to spot drops and prove impact

(13:49) Why you need to apply LLM filters to cut alert noises

(15:06) How to treat session replay as your causal X-ray for true issues

(20:59) Lock down eval metrics before you scale generated UIs

(28:37) Pull instrumentation bugs from sessions and push actionable fixes

(29:43) Pilot session-first pipelines to avoid brittle taxonomies

(31:41) Pair conversational queries for discovery with a GUI for deep analysis

(36:31) Seed adoption by solving internal time-suck problems with copilots

(54:10) Quick tips on how to win at your AI project

—

Where to Find James Evans:

• LinkedIn: https://www.linkedin.com/in/james-evans-7086b3126/

Where to Find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery

Tomasz Tunguz is the founder and General Partner at Theory Ventures. He launched Theory’s debut $230M fund in 2023 and has continued to scale the firm. Previously, Tomasz spent over a decade at Redpoint Ventures, where his blog and analyses helped shape how founders and investors measure SaaS growth. Today, he’s focused on how AI changes go-to-market, moats, metrics, and valuation -  translating old SaaS playbooks into practical advice for AI-native founders.

We recorded this conversation a few months ago, but the lessons are just as relevant today. In this episode, Tomasz explains which SaaS-era rules still apply, what’s changed in an AI-first world (metrics, moats, distribution), and the concrete moves founders and product leaders should make in the next 12–24 months.

--

What you’ll learn:

1. Why SaaS metrics don’t tell the whole story for AI companies, what to track instead.

2. Which kinds of moats matter now (workflow & distribution > pure switching cost).

3. Why distribution is the multiplier for AI products and how to think about embedding AI into workflows.

4. How to prioritize velocity and iteration on model-driven features for compounding advantage.

5. What VCs are now evaluating differently (model costs, workflow embedding, adoption loops).

6. Practical tools and stacks founders can use today to prototype and learn faster.

7. A checklist for founders and operators to prepare for the next fundraising or GTM pivot.

—

What’s covered:

(00:00) Tomasz’s journey: Redpoint to founding Theory Ventures

(06:45) Why AI changes the metrics that matter for early-stage companies

(12:10) Data moats vs. workflow moats: what actually works

(18:35) Why distribution is everything in an AI-first world

(25:42) Lessons from Looker, Dremio, StackRox, and portfolio plays

(32:50) The role of open source and community in AI infrastructure & distribution

(40:15) How VCs are adapting due diligence and their own playbooks with AI

(47:02) The next 24 months — where Tomasz is excited (and cautious)

—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—

Blog: https://tomtunguz.com

X/Twitter: https://x.com/ttunguz

LinkedIn: https://www.linkedin.com/in/tomasztunguz

LinkedIn: https://www.linkedin.com/in/haroonchoudery/

X: https://x.com/haroonchoudery

Elie Schoppik leads Technical Education at Anthropic, where he builds Claude Code’s learning-first workflows and Anthropic Academy to help teams verify outputs, instead of just consuming them.

Before Anthropic, Elie co-founded early ed-tech products and Rhythm School, a Bay Area coding bootcamp that trained thousands of developers. With a background at the intersection of pedagogy and engineering, Elie’s work is focused on designing exercises, evals, and prototype workflows that teach durable judgment and make AI adoption safer and more useful.

In this episode, Elie shares concrete frameworks, playbooks, and hands-on exercises that product and engineering leaders can use to make AI a tutor for their teams, instead of a brittle shortcut.

--

What You’ll learn:

1. Why Claude Code’s learning mode forces people to think with the model, not just copy answers
2. How the Model Context Protocol (MCP) unlocks reliable model ↔︎ data integrations across Gmail, Drive, Slack, and more
3. The end-state prototyping pattern - using screenshots and examples to iterate quickly toward “what good looks like”
4. Why evals (unit tests for LLMs) are critical to avoid brittle deployments and false confidence
5. How to run short, hands-on sprints that beat long training courses for team adoption
6. The self-improving loop: using Claude Code + quizzes to grade, explain, and retrain learning workflows
7. Anthropic Academy’s resources for non-technical and technical learners
8. Elie’s vision for AI as a personalized tutor - adapting to each learner’s style, pace, and goals

--

What's covered:

(02:29) Elie’s background and leading technical education at Anthropic 
(03:41) How AI education differs from coding 
(07:15) Anthropic’s products: Claude Code, MCP, and model actions 
(17:36) Why examples and context matter in prompting 
(18:45) Context engineering: narrowing tasks to get better model output 
(19:59) Core topics to study beyond tool use 
(32:26) Role of education in helping companies join the “successful 5%” 
(35:37) Hackathons as a catalyst for AI adoption in organizations 
(44:59) Dopamine hits, quick wins, and why they matter in AI learning 
(49:03) Can mastering AI tools reduce “AI anxiety”? 
(50:47) Balancing speed vs. responsibility in an AGI timeline 
(52:29) Anthropic Academy resources for technical + non-technical teams 
(55:17) What excites Elie most about the future of AI learning 
(56:03) How parents can prepare kids for an AI-driven future
—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.
—

Where to find Elie Schoppik:

• Twitter/X: https://x.com/eschoppik
• LinkedIn: https://www.linkedin.com/in/eschoppik/
• Anthropic: https://www.anthropic.com

Where to find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery

Jeremy Epling is the Chief Product Officer at Vanta, the leading trust management platform helping 7,000+ companies automate security and prove compliance.

Before Vanta, he spent over a decade building some of the most widely used developer products — from GitHub Actions and Codespaces to OneDrive and Azure. Today, Jeremy is leading Vanta’s AI transformation, rethinking how trust and security can be automated at scale.

—
Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—
(00:00) Jeremy's path from Microsoft and GitHub to leading product at Vanta

(08:05) Why most AI projects fail — and how to pick ones that succeed

(09:48) Data, trust, and how Vanta doesn't train on customer data

(15:10) The messy truth about customer data infrastructure for AI

(25:00) The breakthrough moments that changed everything

(28:42) Voice notes, transcription tools, and how product leaders actually work

(31:26) How AI is killing the traditional PM role (in a good way)

(33:33) Why PMs are shipping code and designers are pushing PRs

(36:05) Show, don't tell — rapid prototyping over endless PRDs

(41:27) How to get marketing in sync with prod

(48:38) Designers owning the frontend (and what that really means)

(51:53) The "boring" features that still matter in an AI world

(57:06) Three things to do before touching any AI project

(59:01) From GitHub Copilot to Vanta — what changed in the AI playbook

(01:01:24) What's next: Vanta's agent is getting scary smart

—
Where to find Jerem Epling:

• Twitter/X: https://x.com/jeremy_epling
• LinkedIn: https://www.linkedin.com/in/jeremy-epling-j40/
• Vanta: www.vanta.com

P.S. Vanta is hiring! If you want to work on AI, security, and trust at scale, check out open roles here: https://www.vanta.com/company/careers

Where to find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery

Des Traynor is the co-founder and Chief Strategy Officer at Intercom, the customer communication platform used by 25,000+ companies worldwide.

He’s spent over a decade helping Intercom scale from a startup into one of SaaS’s most iconic brands. Today, Des is leading the company’s shift into the AI era — rethinking how support, success, and customer engagement work when AI becomes the default.

—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—

(00:00) Introduction — Des's role at Intercom and the company's early AI explorations

(04:25) How to spot the "right first use case" for AI in customer communication

(09:12) Trust, transparency, and shipping AI features that customers actually adopt

(14:48) Balancing speed of iteration with quality when moving AI into production

(19:33) Building an "AI-native" culture inside an established company

(25:07) What incumbents get wrong about adopting AI vs. what startups get right

(32:14) The business impact — how AI is changing Intercom's economics and customer value

(37:56) Lessons from previous platform shifts (mobile, SaaS) applied to AI

(43:20) The future of customer communication with AI at the center

(48:42) Lightning round + closing reflections

—

Where to Find Des:

• Twitter/X: https://x.com/destraynor
• LinkedIn: https://www.linkedin.com/in/destraynor/
• Intercom: intercom.com

Where to find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery

Brandon Sammut is the Chief People Officer at Zapier, one of the earliest and most successful no-code automation companies.

With over 800 fully remote employees, Zapier has been at the forefront of workflow automation for more than a decade. Brandon’s role sits at the intersection of people, culture, and technology — giving him a unique vantage point on how to drive company-wide AI adoption at scale.

—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—

(00:00) Introduction — Brandon's role at Zapier and how AI entered the conversation

(04:12) Why top-down AI mandates rarely work (and what Zapier does instead)

(08:45) Building AI fluency across a remote, global team

(12:30) Zapier's "center of excellence" approach — balancing guidance with experimentation

(17:55) Early wins: how AI started showing up in recruiting, customer support, and marketing

(22:40) The challenge of scale: making AI adoption sustainable vs. shiny

(27:15) Rethinking skills and roles in the AI era — how HR is adapting

(32:02) The cultural playbook: celebrating use cases, sharing learnings, keeping momentum

(38:16) What incumbents get wrong about AI adoption

(42:50) Zapier's bets on where automation + AI converge next

(47:30) Lightning round + closing reflections

—

Where to Find Brandon:

• -LinkedIn: https://www.linkedin.com/in/brandon-sammut-8147b76/ 
• Zapier: zapier.com

Where to find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery

Siqi Chen is the founder and CEO of Runway, a next-gen FP&A -- financial planning and analysis platform.

After experiencing firsthand how broken decision-making was — especially during COVID revenue shocks — Siqi set out to build a tool that empowers operators to make sharper, faster calls. Instead of endless spreadsheets and fragile assumptions, Runway gives teams AI-powered decision support that feels less like Excel and more like an Iron Man suit for operators.

—

Brought to you by Autoskills (https://autoskills.ai) – Helping teams go from AI-curious to AI-ready with tailored workshops and hands-on adoption.

—

00:00 Introduction — Siqi's background and how Runway began

05:03 Ambient AI in practice — embedding intelligence subtly across workflows

09:36 Scaling AI judgment: how Runway spreads AI "taste" through hiring, culture, and team norms

13:15 From hackathons → hack sprints: why rapid iteration is key, and how to turn prototypes into production

17:10 Avoiding AI sprawl — letting teams experiment without getting lost in tools

19:24 Practical wins: how Runway transformed lead scoring, customer success, recruiting, and internal workflows with AI

25:40 How Siqi keeps headcount less than 100 using AI

33:15 How to empower non-technical teams to supercharge their workflows with AI

43:22 Tools Runway uses internally

47:40 Lessons learned from AI adoption sprints — what sticks vs. what fizzles

51:05 The future of AI in decision-making — where FP&A is headed

55:18 Lightning round + final thoughts

—

Where to Find Siqi:

• Twitter/X: https://x.com/blader
• LinkedIn: https://www.linkedin.com/in/siqic/
• Runway: runway.com

Where to find Haroon Choudery:

• LinkedIn: https://www.linkedin.com/in/haroonchoudery/
• X: https://x.com/haroonchoudery