Code Riff

Bryan is an oil market analyst. Every morning he spends one to two hours reading news - Reuters, CNN, CNBC - trying to pull out the actual facts from opinion and spin. Two articles about the same event sometimes say opposite things.

We tried building him an AI tool using Claude Code. It scraped hundreds of articles, stripped out the editorializing, and pulled just the verifiable facts. It covered about 80-90% of what he needs. 

What you'll hear in this episode:
- An oil analyst explaining the daily problem: separating fact from spin in news headlines
- Building a fact extraction tool with Claude Code, live in one session
- The moment two articles about the Strait of Hormuz directly contradicted each other
- Why "dangerously skip permissions" exists in Claude Code (cowboy mode)
- Bryan's live feedback: what worked, what didn't, and what he'd actually use
- Yaohong on why confidence scores without defined criteria will hallucinate

Hosts: Eric Tan (non-technical builder) & Yaohong Ch'ng (engineer, Superuser HQ, ex-Stashaway head of Data)
Guest: Bryan - oil market analyst

Got a problem you want us to solve live? Fill out the form:
https://forms.gle/DSyLzPAoR6x2M4Np9

Connect with us:
- Eric on LinkedIn: https://www.linkedin.com/in/erictisme/
- Yaohong on LinkedIn: https://www.linkedin.com/in/yaohongchng/
- WhatsApp community: https://chat.whatsapp.com/Dmp5eEEsAZhJTB6LjcIG3c?mode=gi_t
- Substack: https://substack.com/@coderiff
- Email: [email protected]

Tools used in this episode:
- Claude Code: https://claude.ai/code
- Gemini API (news classification/analysis): https://ai.google.dev/
- Python (RSS scraping + fact extraction)

Tools we use:
- Buzzsprout (podcast hosting): https://www.buzzsprout.com/?referrer_id=2371679
- Snipd (AI podcast highlights): https://get.snipd.com/pAbF/36jzrvki

LEARN ALONG - Glossary:
- RSS feeds: A way to automatically pull articles from news sites without visiting each one. Like a subscription pipe that delivers headlines to your tool.
- API keys: Password-like codes that let your tool access paid services like AI models. You keep them secret and they track your usage.
- Hallucination: When AI confidently generates information that sounds right but isn't real. The biggest risk when you need verified facts for trading decisions.
- Sandbox: A restricted environment that limits what code can do on your computer. A safety net when running AI-generated code.
- Dangerously skip permissions: A Claude Code setting that removes the safety sandbox. Useful for speed, risky if you don't trust what the code is doing.
- Scenario analysis: A structured way of thinking through "if X happens, then Y follows." Bryan uses a specific framework: impact, market response, price implications.

ABOUT:
Code Riff - messy real-world problems, solved with AI, so you can too.

One of us can't code. The other's been coding since he was 13. Every episode, someone brings us a real problem and we try to solve it with AI in one hour live.

Chuyi quit her corporate job to help manage a wet market vegetable store at Marine Terrace. She tracks competitor prices and supplier orders through WhatsApp and spreadsheets — all by hand.

So we used Claude Cowork to build her a competitor pricing guide and supplier management dashboard in one hour, having fun while doing it together.

What you'll hear in this episode:
- A wet market store operator walking through her daily pricing and supplier workflow
- Using Claude Cowork to scrape and compare prices across FairPrice, Sheng Siong, Cold Storage, and more
- Building a supplier management tracker with price logging, margin calculations, and trend analysis
- AI "hallucinating" wet market prices — and Yaohong catching it live
- Honest conversation about scraping ethics, data accuracy, and why "your data will never be clean"
- Why not everything needs code — sometimes a spreadsheet is the real unlock

Hosts: Eric Tan (non-technical builder) & Yaohong Ch'ng (engineer, Superuser HQ, ex-Stashaway head of Data)
Guest: Chuyi (Le Fresco, 50 Marine Terrace)

Got a problem you want us to solve live? Fill out the form:
https://forms.gle/DSyLzPAoR6x2M4Np9

Connect with us:
- Eric on LinkedIn: https://www.linkedin.com/in/erictisme/
- Yaohong on LinkedIn: https://www.linkedin.com/in/yaohongchng/
- WhatsApp community: https://chat.whatsapp.com/Dmp5eEEsAZhJTB6LjcIG3c?mode=gi_t
- Substack: https://substack.com/@coderiff
- Email: [email protected]

Tools used in this episode:
- Claude Cowork: https://claude.ai
- Google Sheets: https://sheets.google.com

Tools we use:
- Buzzsprout (podcast hosting): https://www.buzzsprout.com/?referrer_id=2371679
- Snipd (AI podcast highlights): https://get.snipd.com/pAbF/36jzrvki

Also discussed: Notion, Airtable, Google CLI, Gemini

LEARN ALONG — Glossary:
- Claude Cowork: A browser-based AI assistant from Anthropic that can control your files and browse the web. Costs ~$30 SGD/month on the Pro plan.
- Scraping: Automatically collecting data from websites using a bot instead of copying it by hand. Some websites block this.
- Tokens: The units AI models use to process text. More tokens = more cost. Like minutes on a phone plan.
- Hallucination: When AI confidently makes up information that isn't real. In this episode, it estimated wet market prices at "20-40% cheaper than supermarkets" — a guess, not data.
- MCP (Model Context Protocol): A way for AI agents to connect to external tools like Notion or Google Sheets. Think of it as a universal adapter plug for AI.
- Network egress: A security setting that controls which websites an AI tool can access from your computer. Like a bouncer checking IDs.

ABOUT:
Code Riff — messy real-world problems, solved with AI, so you can too.

One of us can't code. The other's been coding since he was 13. Every episode, someone brings us a real problem and we try to solve it with AI in one hour live.

An ESG analyst at a bank spends her morning commutes reading news across Reuters, ESG News, and Financial Times, etc, then manually logging it into Excel and Word. She tried ChatGPT but got hallucinated headlines.

So we automated the whole thing in one hour using Claude Code from scratch, having fun while doing it together.

What you'll hear in this episode:
- A real sustainable finance analyst walking through her daily news workflow
- Building a Python news scraper live with little to no prior coding experience
- AI classifying articles into sustainable finance, policy, and country news
- Setting up a Telegram bot that delivers a formatted Word document daily
- Honest explanations of API keys, prompt injection, and context rot
- A conversation about AI in banking - augmenting jobs, not replacing them

Hosts: Eric Tan (non-technical builder) & Yaohong Ch'ng (engineer, Superuser HQ, ex-Stashaway head of Data)
Guest: Wachel - Sustainable Finance Analyst

Got a problem you want us to solve live? Fill out the form: https://forms.gle/DSyLzPAoR6x2M4Np9

Subscribe to our newsletter: https://substack.com/@coderiff?utm_campaign=profile&utm_medium=profile-page

Tools from this episode:
- Claude Code: https://claude.ai
- Newsdata.io: https://newsdata.io
- Telegram (BotFather): https://t.me/BotFather

Connect with us:
- Eric: https://www.linkedin.com/in/erictisme/
- Yaohong: https://www.linkedin.com/in/yaohongchng/
- Superuser HQ: https://superuserhq.com/
- Email: [email protected]

Code Riff - messy real-world problems, solved with AI, so you can too.

I don't write code. Yaohong's been coding since he was 13. We sat down to figure out why we're making a podcast about building things with AI - live, on camera, with real people's real problems.

In this episode, we talk about:
- Why this moment is important for non-coders to build with AI
- Singapore's 2026 budget and AI adoption
- The problem with AI consultants selling to SMEs who don't understand AI
- Why we picked Claude Code over Lovable, Bolt, and Cursor (for now)
- Pains of 24/7 OpenClaw bots
- What guests / subscribers can expect

Hosts: Eric Tan (non-technical builder) & Yaohong Ch'ng (engineer, Superuser HQ)

Got a problem you'd love to solve with AI? Submit it here - we'll try to build it live:
https://forms.gle/DSyLzPAoR6x2M4Np9

Timestamps:
00:00 - Trailer
00:43 - Why we're doing this
01:59 - When AI became real for us
04:25 - How we met and what's changed since
05:16 - Building an AI-native company
07:22 - Why now
08:23 - Yaohong's backstory
10:40 - Our format and a little rant on OpenClaw bots
14:49 - Singapore's 2026 budget and AI strategy
16:54 - The problem with AI consultants
18:36 - What we're really trying to do
20:15 - Limits of AI
23:19 - Why Claude Code vs other vibecoding tools
26:17 - Using AI for writing
28:38 - Kids and AI
30:42 - Risk of AI dependence
35:51 - What guests can expect

Links:
Submit your problem: https://forms.gle/DSyLzPAoR6x2M4Np9
Superuser HQ: https://superuserhq.com/
OpenClaw (AI agent framework): https://github.com/openclaw/openclaw
Eric's LinkedIn: https://www.linkedin.com/in/erictisme/
Yaohong's LinkedIn: https://www.linkedin.com/in/yaohong/

Learn along (tech terms from this episode):
- Claude Code: An AI tool you talk to in your terminal. You describe what you want in plain English, it writes and runs the code.
- Vibecoding: Building software by describing what you want, not writing code yourself. What Eric does.
- Terminal / command line: The text-based interface where you type commands. Like texting your computer instead of clicking buttons.
- API key: A password that lets your app connect to another service. The "annoying part of vibecoding" according to Eric.
- Lovable / Bolt / Replit: Web-based AI app builders. Simpler than Claude Code but less flexible.
- Cursor: An AI code editor. More visual than Claude Code, less powerful.
- OpenClaw: An open-source AI agent framework. Yaohong's team built personal virtual assistants (Ram and Connie) that collaborate with each other in Slack channels. Think AI coworkers that can switch roles and work on projects together.
- Supabase: A database service where apps store data. Like a spreadsheet in the cloud.
- Sycophancy: When AI agrees with everything you say instead of pushing back. Can lead people down dangerous rabbit holes.

We get on a call with someone, hear what's bugging them, and try to solve it with AI in an hour. That's basically the whole show.