Bits, Bytes and Business

# Show Notes: Vibe Coding and Real World Problems: Why AI Still Needs a CTO

**Guest:** Anuraag Jain, CTO, Executive Coach and Author

**Website:** [growthvariable.com](https://www.growthvariable.com)

**LinkedIn:** [Anuraag Jain](https://www.linkedin.com/in/anuraagjain)

---

## About This Episode

Anuraag Jain is a 30-year tech industry veteran who started as an Oracle DBA in India without an engineering degree and became a CTO in under nine years. He was the first developer hired at Betfair, where he built the betting exchange engine behind one of the era's biggest fintech IPOs, and served as CTO at The Hut Group (THG) during their run as the UK's fastest-growing tech company.

In this episode, Bhairav and Anuraag talk honestly about vibe coding: what it can do, where it breaks down, and why thirty years of building things under pressure still matters in an AI-assisted world.

---

## What We Cover

- How two non-traditional hires found their way into tech and why that background shaped how they build

- The early days at Betfair: manual bet settlement, shoestring infrastructure, and recovering 36 hours of betting data from server log files at 2am

- What scaling really means when you have no safety net

- The 80/20 rule for vibe coding: what AI handles well and what still requires human judgment

- Security blind spots in AI-generated code and why even experienced engineers get caught out

- The hidden cost of running AI tooling against a large code base

- Repo to Pitch: the open source tool that turns a code base into a full pitch deck and documentation suite

- RepoWatch: how to track exactly what AI has changed in your code, commit by commit

- *Prompt Engineering the Subconscious*: the book that connects AI prompting to mindset and reached the UK top 50

- What the fractional CTO market actually looks like right now, and the three types operating in it

- Practical advice for anyone trying to break into tech or move toward a CTO role today

---

## Key Takeaways

- AI can handle roughly 80% of what used to require a developer. The remaining 20% is the domain logic that requires someone who actually understands the problem.

- The AI does not know which parts of your code cannot be touched. Unless you tell it, it will modify anything, and you may not notice until something breaks.

- Vibe coding a product does not make it production-ready. Security, scalability, and edge-case logic still need human attention.

- Building something real, however small, is still the best way to develop the judgment the CTO role requires. The tooling available now means there is no excuse not to.

---

## Links

- Anuraag Jain on LinkedIn: [linkedin.com/in/anuraagjain](https://www.linkedin.com/in/anuraagjain)

- Growth Variable: [growthvariable.com](https://www.growthvariable.com)

- Repo to Pitch (open source): [growthvariable.com/repotopitch.html](https://www.growthvariable.com/repotopitch.html)

- RepoWatch: [growthvariable.com](https://www.growthvariable.com)

- *Prompt Engineering the Subconscious*: [book.growthvariable.com](https://book.growthvariable.com)

---

## About the Atom CTO Podcast

The Atom CTO Podcast, transitioning to Bits, Bytes and Business, features conversations with founders, CTOs and entrepreneurs about building technology businesses in the real world. Hosted by Bhairav Patel.

Subscribe wherever you get your podcasts.

This week veteran tech CTO Anuraag Jain, author of Prompt Engineering the Subconscious is answering 5 questions: https://book.growthvariable.com/ i

James found me online looking for fractional CTO services. That conversation eventually made me an investor. In this episode I sit down with all three co-founders of Koloso — James, Petra and Sowi — to talk about building a meaningful EdTech product in Zambia with limited capital, no technical background, and a problem that affects millions of children across Sub-Saharan Africa.

**The problem:** Zambian classrooms regularly have 60 to 100 children and one teacher. Teachers have no reliable way to track who is keeping up or falling behind, and a mountain of admin on top. The slogans around leaving no child behind simply don't translate into classroom reality.

**The product:** Children play a two-minute, ten-question quiz aligned to the Zambian national curriculum. Teachers get a real-time dashboard showing individual progress over time. Parents receive practical tips for reinforcing learning at home, in everyday Zambian contexts. Version two covers the entire teaching loop, from lesson planning through to compliance reporting.

**The challenges:** Schools are cautious adopters. The "lazy teacher" objection was real and unexpected. Annual procurement cycles create brutal seasonal gaps. Building without technical founders nearly finished them.

**AI at the right moment:** Koloso started using AI-assisted development around eight to ten months before this conversation. Problems that once required lengthy specifications now get resolved in half an hour. But if AI had arrived two years earlier they would have built the wrong product. A year later and they would have run out of money. The timing was critical.

**An African solution to an African problem:** Khan Academy hasn't gained traction in Zambia because it isn't aligned to the Zambian curriculum. Koloso's questions use Zambian names, Zambian currency, and Zambian contexts. The family-centred learning philosophy reflects Zambian cultural values in a way that imported solutions don't.

---

## Timestamps

| Time | Topic |

|------|-------|

| 00:32 | How the founders met and their backgrounds |

| 04:00 | Why data is the core problem in Zambian classrooms |

| 07:00 | Building the first prototype |

| 11:10 | How the app works |

| 13:58 | School adoption challenges |

| 17:28 | African solution to an African problem |

| 26:53 | AI-assisted development and why timing mattered |

| 35:01 | Advice to their younger selves |

| 37:21 | Where to find Koloso |

---

Koloso is live in Zambia, South Africa, Nigeria and Uganda, with Malawi in progress. Available on iOS and Google Play. Pre-seed round of $250,000 currently open. Actively seeking investors and schools.

**www.koloso.app**

*Subscribe to the Atom CTO Podcast for more conversations with founders building things that matter.*

This week I'm joined by Sowi, Petra and James, the founders of Koloso, a Zambian EdTech Startup

We're back! A new season of the Atom CTO podcast will be coming to your ears over the coming weeks so stay tuned!

What used to be relatively straightforward—“Twitter for debate, Facebook for fun, Instagram for cool visuals, LinkedIn for business”—has become chaotic, politicized, and in many ways more dangerous for founders and brands.

In this episode with Colin Kelly we explore:

-Why the old rules of social media are gone and why nobody can honestly claim to “know” how it all works anymore.

-How social platforms have shifted from organic discovery engines to paid advertising systems and what that means for startups.

-The renewed importance of old‑school sales and marketing (phone calls, events, letters) and using social media as supporting evidence rather than your primary growth engine.

-The tension around brand values, politics, and polarisation, and why being loudly “values‑driven” online can now backfire.

-How to show your human side online without falling into inauthentic “I’m crushing it” performance.

-Where AI, SEO, and search are heading, including a fascinating discovery about Instagram posts appearing in Google’s AI Overviews.

In this episode, Bhairav and Alan bring in AI practitioner Sanjay Rakshit to unpack what’s really going on with artificial intelligence – beyond the noise, hype, and fear.

They explore what AI actually is (and isn’t), why no one is truly an “AI expert” yet, and how founders and business leaders should think about AI as a power tool rather than a magic replacement for people.

The conversation ranges from:

- The history and evolution of AI and why it suddenly “caught fire” with ChatGPT.

- How hype distorts reality for founders, investors, and even universities.

- How to differentiate between genuine practitioners and people who are just one chapter ahead of you.

- Why AI augments good practitioners instead of replacing them – and what that means for developers, doctors, and small businesses.

By the end, you’ll be less intimidated by AI, more skeptical of big promises, and clearer on how to use it intelligently in your business.

In this week's podcast, I'm talking with Kiley Tan, a co-founder and legal expert in AI with The Legal Director, who sheds light on how startups and businesses can navigate the complex landscape of AI usage, focusing on legal risks, copyright issues, and data protection.

Key Discussion Points:

Introduction to AI in Business: Kiley Tan shares his journey into AI, explaining how his curiosity led him to pursue an AI course to better understand its applications in business.

Legal Concerns in AI: Discussion on common legal challenges related to AI, including data protection, copyright issues, and ownership of AI-generated content.

Using Internal Data: Importance of leveraging proprietary internal data which can provide a significant advantage as it's owned by the company.

The Role of AI in Legal Drafting: Examination of AI tools in drafting legal documents and the complexities involved.

Agentic AI Risks: Examination of risks associated with autonomous agents, including liability issues if AI actions lead to unintended consequences.

For more info on The Legal Director, follow this link: https://www.thelegaldirector.co.uk/

You can see Kiley's profile here: https://www.thelegaldirector.co.uk/team/kiley-tan/

You can follow Kiley here: https://www.linkedin.com/in/kileytan/

This week's guest is Kiley Tan, Consultant Solicitor specialising in AI with The Legal Director

You will learn the following:

- There are patterns and challenges that companies commonly face at different stages of their growth. These challenges are not random.

- Building a strong foundation and getting the basics right is essential for success in business.

- Hiring the right people and aligning their values with the company’s values is crucial, especially when making the first few hires.

- Conflict resolution can vary from case to case, but there are usually patterns that can be recognized. Choosing the path that minimizes resentment and disappointment is often the best approach.

- Businesses can be rated on a scale of one to ten, and understanding the size and shape of a business can help determine the type of experience and skills needed for hiring and growth.