Follow me to learn how you can leverage AI to boost your productivity and accelerate your career. Scaled products to 8 Million+ users and built the world's biggest AI newsletter with 1 Million+ readers.

Zain Kahn

This guy from Cambridge plays with Claude Code like he owns it.

He talks about his workflow in 2026 and here's what's new: 

1. Video-Based Specification (Spec Phase)

> Screen Recording: 

Instead of writing a spec from scratch, find an existing product that is similar to your idea. Record your screen while using it and talking through your specific feature ideas and changes.

> Generate PRD: 

Upload this video to Gemini 1.5 Pro (referred to as Gemini 3 Pro/Free Pro in the video context) and ask it to generate a Product Requirement Document (PRD).

> Refine Spec: 

Use the "Ask User Question" tool in Claude Code. Prompt it to interview you about the generated spec to fill in missing details (e.g., "How should the emoji picker be positioned?").

> Package Discovery: 

Feed the refined spec into ChatGPT with "Heavy Thinking" (likely OpenAI's o1/o3 models) to search for and recommend specific, well-maintained GitHub packages (e.g., for a WYSIWYG editor) to avoid building complex components from scratch.

2. The Orchestrator Role

> Design Feedback Loops: 

Your primary job is not to write code but to design loops where the agent can build, fail, and learn.

> Monitor & Update: 

Watch the agent’s reasoning. If it makes a mistake, don't just fix the code—update the claude. md (project instructions file) to prevent that specific mistake from happening again.

> High-Level Decisions: 

Make the architectural decisions the agent can't, such as choosing the database or specific tools.

3. Model Choice & Tools

> Model Selection: 

Use Opus 4.5 for building large-scale features and GPT-5.2 for architecture and debugging (specific future models mentioned in his 2026 context).

> Voice Dictation: 

Use HyperWhisper to dictate prompts significantly faster than typing.

4. Execution & "Parallel Vibe Coding"

> Parallel Agents: 

For small, well-defined tasks (like extracting hard-coded strings for translations), spin up multiple sub-agents to work on different parts of the project simultaneously.

> Avoid Conflicts: 

Do not use parallel agents for large features within the same project to avoid complex merge conflicts ("meshing issues").

> Sub-agents for Multi-Project Fixes: 

If a bug exists in a template used by multiple projects, spin up one sub-agent per project to fix them all in parallel.

5. Review & Maintenance

> Planning Mode: 

Use Claude Code’s "Planning Mode" to prevent Architectural Drift, ensuring the agent sticks to the original design vision over time.

> Shape of Diffs: 

Inspect the "shape" of the code changes (diffs) to ensure they are manageable and align with expectations before accepting them.

> Forking for Learning: 

If the agent does something surprising or complex, fork the session. In the forked session, ask "Why did you do that?" or request diagrams to understand the code without polluting the context of the main working session.

Get more tutorials here: https://lnkd.in/e64Jvdrt

Follow me to learn how you can leverage AI to boost your productivity and accelerate your career. Scaled products to 10 Million+ users.

Traditional computer science degrees can't keep up with this pace.

Here are the 15 engineering blogs that actually teach real-world scale:

1 → THE CODE
Scale at global level.
Teaches you how to code with AI and AI Agents
https://lnkd.in/e64Jvdrt

2 → Airbnb Tech Blog
Product engineering done right.
User focused systems.
https://airbnb.tech/blog/

3 → Meta Engineering
Billions of users infrastructure.
Performance at extreme scale.
https://engineering.fb.com

4 → AWS Architecture
Cloud design patterns revealed.
Production ready blueprints.
https://lnkd.in/gqjh88PS

5 → Netflix TechBlog
Streaming at massive scale.
Chaos engineering explained.
https://lnkd.in/gNE8Cs6Z

6 → Discord Engineering
Real time systems mastery.
Handling millions of connections.
https://lnkd.in/gea6kSwt

7 → Google Research
Cutting edge tech breakthroughs.
Tomorrow's solutions today.
https://lnkd.in/gjzDu8wp

8 → NVIDIA Developer
GPU powered computing insights.
AI infrastructure secrets.
https://lnkd.in/dt_BW7qF

9 → Slack Engineering
Collaboration platform internals.
Distributed systems at scale.
https://slack.engineering

10 → Cloudflare Blog
Internet infrastructure deep dives.
Security and performance wins.
https://lnkd.in/gqa-HE6h

11 → Figma Tech Blog
Collaborative design systems.
Browser performance magic.
https://lnkd.in/g2KF_eX9

12 → Shopify Engineering
E-commerce at scale.
Black Friday traffic handled.
https://lnkd.in/g7DVzRAc

13 → Stripe Engineering
Payment systems that work.
Financial infrastructure explained.
https://lnkd.in/gfMr-jEr

14 → Microsoft Engineering
Enterprise scale solutions.
Developer tools insights.
https://lnkd.in/g3Mi8XES

15 → GitHub Engineering
Code collaboration at scale.
Developer platform secrets.
https://lnkd.in/gPHaD6AC

16 → Bonus
System design deep dives.
https://lnkd.in/g-tP7eEc

Save this for later.
Repost ♻️ for aspiring engineers.

Follow me for more AI playbooks and engineering resources.

Claude Code is only as good as your orchestration workflow.

I spent 6 months testing it so you don't have to. 

[ P.S. You can get my Ultimate Claude Code guide for engineers here at no cost: https://lnkd.in/e64Jvdrt ]

So Boris Cherny (the creator of Claude Code at Anthropic) recently shared the internal best practices his team actually uses daily. 

Someone brilliantly distilled those threads into a structured CLAUDE .md file you can drop straight into any project root.

It acts as a system prompt. Turns Claude into a far more autonomous, rigorous engineering partner.

Here's what it does:

→ 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄 𝗢𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻: Mandates "Plan Node Default" for any task over 3 steps. Uses subagents liberally to keep the main context window clean.

→ 𝗦𝗲𝗹𝗳-𝗜𝗺𝗽𝗿𝗼𝘃𝗲𝗺𝗲𝗻𝘁 𝗟𝗼𝗼𝗽: This is the real magic. After ANY correction, it updates a tasks/lessons .md file. You're building a compounding system where the mistake rate drops over time because it actively learns from your feedback.

→ 𝗩𝗲𝗿𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗕𝗲𝗳𝗼𝗿𝗲 𝗗𝗼𝗻𝗲: Can't mark a task complete without proving it works. Diffs behavior. Runs tests. Checks logs. The bar? "Would a staff engineer approve this?"

→ 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗕𝘂𝗴 𝗙𝗶𝘅𝗶𝗻𝗴: Zero hand-holding. Point it at failing CI tests or error logs... it just goes to work. No constant context switching from you.

→ 𝗦𝘁𝗿𝗶𝗰𝘁 𝗧𝗮𝘀𝗸 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁: Forces a "Plan First" approach written to a todo .md with checkable items before any implementation starts.

And perhaps the most important part?

It forces the AI to prioritize simplicity, find root causes instead of temporary fixes, and minimize the blast radius of every change. Senior developer standards. Not shortcuts.

If you're spending hours a day in the terminal with AI, setting up a strong .md instruction file like this isn't optional anymore. It's the difference between AI that drifts and AI that compounds.

It takes time to set up. But if you do, you're ahead of almost everyone else.

Lemon is the most powerful tool I've used since ChatGPT.

You literally say things with your voice and it does it for you instantly.

Pure magic.

Founder @ Growth Marketing Agency - Single Grain, Podcaster @ Marketing School + Leveling Up

Single Grain is a digital marketing agency focused on driving scalable and predictable revenue growth using Facebook advertising, Google Ads, YouTube advertising, SEO, video production and content marketing. 

We've worked with companies such as Amazon, Uber, Nordstrom and venture-backed startups to create and execute on custom-tailored growth strategies.






-> For those of you that are interesting in marketing and personal growth tips from top entrepreneurs, please visit http://www.levelingup.com for our free podcast. You'll hear from billionaires, authors, and other high performers.

-> For my daily marketing podcast, go listen to Marketing School at https://www.singlegrain.com/marketing-school-podcast



Skills:
SEO
Paid Search (Pay per click advertising)
Affiliate Marketing
Display Advertising
Mobile Advertising
Remarketing/Retargeting
Facebook Marketing
Video Advertising
Twitter Marketing
Google Analytics
Conversion Rate Optimization
Sales
Strategic Partnerships
Content Strategy/Development
Business Development
Social Media
Community Development
Operations
Affiliate marketing
Affiliate program management
Retargeting
Growth Hacking

Eric Siu ✓

When I took over Single Grain, it was a failing SEO shop.

Revenue was tanking. And I knew one bad hire would absolutely destroy what little we had left.

That's why I'm partnering with LinkedIn.

I've learned after 15 years of building agencies: you don't need more resumes. You need the right people. Fast.

What I like about #LinkedInHiringPro is that I can describe what I actually need in my own words. The AI agent then surfaces and invites qualified candidates I probably wouldn't have found on my own.

Scaling a business is hard. Hiring the right person doesn't have to be.

Learn how LinkedIn Hiring Pro helps you hire right the first time. Get $100 off your first sponsored job post — linkedin.com/EricS. (Terms apply) 

#LinkedInPartner

If you are obsessed with Claude Code, OpenClaw, etc., we want to hire you.

First step is to go through our ‘beat claude’ challenge at Single Grain | Start Growing . 🙂

I believe that this will be the future of hiring. Feel free to take the template for your own company.

Link in first comment.

No alternative text description for this image

84% of the world has never used AI.

Let that satisfying visual sink in for a second.
Each dot in this chart represents ~3.2 million people. 2,500 dots = 8.1 billion humans.

The gray? That’s ~6.8 billion people who have never touched an AI tool. Not once.

The green at the bottom? ~1.3 billion free chatbot users. That’s only 16%.

The tiny sliver of yellow? ~15-25 million people paying $20/month for AI. That’s 0.3%.

And the red dots you can barely see? ~2-5 million people using coding scaffolds like Claude Code or Cursor. 0.04% of the planet.

I run a business where AI fluency with Claude Code is now mandatory. We’ve replaced significant headcount costs with autonomous agent systems. We build business tools with Claude Code.

Most founders I talk to think AI adoption is “everywhere.” It’s not even close.

If you’re reading this, you’re likely in the top 1% of AI users on Earth. The opportunity gap between those who use AI and those who don’t is only getting wider.

We’re at the dial-up era of AI. The people building skills now will compound their advantage for the next decade.

What’s your take on the adoption curve?

Many people talk about OpenClaw for personal use cases, but not enough for business use cases.

We used it to book a meeting with a multi-trillion dollar company.

I'm going to share a few business examples that will hopefully give you ideas. I think a lot of people are on the fence right now because they're either scared, lazy, or both.

Once you give it a shot, you'll become addicted.

And if you're non-technical, don't worry, because I'm in the same boat.
OpenClaw is an open-source autonomous agent you can run locally or on a server. Almost anything you can do on a computer, it can do.

Here are 4 real business use cases:

Daily Deal Finder

OpenClaw goes through our CRM, sales transcripts, and based on our goals, figures out the right people to talk to and what the angle is. This alone booked me a call with a trillion (with a T) dollar company. On a conversation we forgot about, it quickly gave me the context, angle, and outreach message.

Ad Creative Scaler

A prospect recently said they wanted 5,000 AI creatives per month. I had the agent look through their ad library, website, understand their brand, and come up with UGC ad concepts that scale. It even created quick video and photo mockups using Gemini's latest models.

Social Media Clipper

I record 80 minutes a week for the Marketing School podcast and we cover a lot of topics that often get missed. That's a lot of wasted potential. My social bot now finds and clips the best moments.

SEO/AEO Agent

Having an agent feed me insights is nice, but let's be real. We all want something to just do the work for us. I had it scan my site for opportunities (connected to GA4, GSC, and Ahrefs API), create production-grade content using ClickFlow, and place the drafts directly inside my CMS.

How This Scales:

My primary "chief of staff" is Alfred. He manages my growing squad of agents: Oracle (SEO), Flash (Social), Green Arrow (Sales), and more. Each agent earns autonomy over time.

New agents start restricted. Insights and data reports only. As they prove reliable, we unlock execution rights for drafting, outreach, purchasing.

Keep a human in the loop for high-stakes decisions.

Who else is playing with OpenClaw right now? And if not, why not?

Join Neil Patel & Eric Siu live as they talk about the latest in AI and marketing

taplio

The Battle

Zain Kahn

Eric Siu ✓

Followers

Posts

Engagements

Likes

Performance ScoreZain Kahn's LinkedIn Performance

Performance ScoreEric Siu ✓'s LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Zain Kahn

Eric Siu ✓

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Zain Kahn's LinkedIn Performance

Performance ScoreEric Siu ✓'s LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreZain Kahn's LinkedIn Performance