Chill Build: Automation & Traffic
Want to back the stream? The «داعم» membership on YouTube is EGP 60 a month: badges, emoji, and I answer your questions first on stream. Become a supporter
The gist
The stream centers on building a controllable automation system that organizes AI agents as a graph and turns research into content, short clips, and multi-platform posts, supported by a calendar, analytics, reports, and Telegram-based coordination. Practical work included checking the publishing pipeline, correcting topic relevance, refining tone and calls to action, and dividing reasoning tasks from deterministic local scripts. Luqman repeatedly emphasized that quality, measurement, and human review should precede scaling or unattended operation. In his view, the larger challenge is no longer software construction alone but acquiring and measuring traffic through tools such as Search Console and Google Analytics, then improving the product and pricing once enough evidence exists. Audience questions covered coding tools and plans, Codex versus Claude, learning through Micro-SaaS, agent tools, e-commerce platforms, SEO, editing, and AI video generation. Other discussions covered opinions on AI safety and the software industry’s future, beginner and freelancing advice, creator-platform monetization, payment-platform choices, the risk of exposed API keys, and the ethics of AI-altered product images. Claims about unidentified models, agent security, usage limits, and tool costs were treated cautiously rather than as established facts. The automation, publishing, and traffic outcomes remained unresolved: some tests and refinements looked promising, but several workflows stalled or were still being configured, and there was not yet enough performance data to call the system successful.
131 moments from the stream
- 3:10:09
Results cannot be evaluated before generating traffic
Answering a question about an in-development project, the speaker explained that there was not yet enough traffic or impressions to judge results. The stated plan was to build organic traffic first, potentially try advertising later, and optimize other areas once traffic arrives.
- 3:11:54
Morning briefing and discussion of the CEO role
The speaker referred to preparing a morning briefing as part of something transcribed as ‘Omar,’ then asked to discuss the new CEO’s role in more detail. The short excerpt does not clarify the referenced system or implementation.
- 3:12:52
Q4 timing as a reason for urgency
The speaker said there was generally no need to rush results, except that the approaching Q4 period could justify greater urgency. This was a time-specific observation from the recording, not a current calendar claim.
- 3:15:19
Could AI Become Accessible to Everyone?
The speaker offers a personal prediction that deeper adoption of AI in working life could reduce its cost and broaden access over the coming years, comparing its possible spread across social groups to mobile phones.
- 3:17:33
Improving Automated Posts with Clear Calls to Action
While reviewing automated post output, the speaker notes a substantial improvement after small changes but says important pieces remain missing. A key recommendation is a rule requiring most posts to include a clear call to action at the beginning, end, or both.
- 3:19:20
What Is an IPO, and Could It Affect AI Service Pricing?
The speaker explains in general terms that an IPO makes shares of a formerly private company available to public investors. He then offers an explicitly uncertain personal prediction that new capital following a possible AI-company listing could support investment and changes in pricing or efficiency.
- 3:19:52
Trustworthiness of AI-Generated Product Images
The speaker discusses trust in e-commerce advertising that uses AI to alter product images. In his view, showing an item in a way that does not match what the customer will receive can be misleading, even when the change appears to be a visual enhancement; an exact representation would be a different case.
- 3:22:58
Choosing Between Default and Extended Context Windows
The speaker says an extended context window is available but that the default works well for him and may have been chosen deliberately for design reasons. He notes that users prefer different settings and that he is accustomed to both, without claiming one is universally superior.
- 3:23:49
Being Cautious When Judging Models You Rarely Use
Asked about a model whose name is unclear in the captions, the speaker avoids offering a broad assessment because he rarely uses Chinese models now. He says his current experience is concentrated mainly on Codex and Claude, so his view of alternatives is limited.
- 3:25:21
Building Automation Through Iterative Improvement
The speaker explains that automated output is not fully accurate at first, and that correcting errors and tuning individual elements over time may reduce the need for manual review.
- 3:25:53
Audience Note on a Codex-Like Alternative
An audience member mentioned an unclearly named Codex-like tool, claiming it offered a limited free trial, produced good results, and used fewer resources; this was an audience observation, not a documented test on the page.
- 3:29:22
Choosing a Call to Action for a Marketing Video
The review considers adding a call to action leading either to a Shopify app installation or a website scan, alongside a positive assessment of the target video quality.
- 3:32:46
Mapping Content Production and Traffic Sources
The speaker requests an org-chart-style hierarchy showing daily posting volume and the workflow across connected teams, then links that map to a plan expected to generate most traffic, presented as an expectation rather than a proven result.
- 3:34:26
Database and Hosting for a SaaS Application
The speaker clarified that he does not use Hostinger to host the application being discussed, and uses Supabase for its database and backend alongside other tools.
- 3:40:26
Reducing Errors in Product Video Generation
The speaker evaluated a format that shows one product from several angles and considered it less prone to generation errors than products with geometry-sensitive details such as phones. He then requested a hierarchy showing how many posts each department would generate per day.
- 3:45:57
Automated Publishing Plans and Performance Reports
The speaker discussed calculating daily and monthly video volume and distributing it across platforms including Instagram Reels, Facebook Reels, YouTube Shorts, LinkedIn, and X. He also described a recurring report summarizing team successes, failures, and progress toward broad goals such as traffic sources, installs, and conversions, with the reporting cadence still open to adjustment.
- 3:49:04
Quality Control and Tone for Automated Personal Content
The speaker distinguished company-branded content from content published under his personal name, arguing that personal posts on LinkedIn and X require a higher quality bar and review before publication. He discussed revising saved instructions to define a more natural personality and tone so generated copy appears human-written, alongside an example of an automated promotional video with voice-over and music.
- 3:54:36
Checking Posting-Cadence Logic
The speaker questions a mismatch between an expected daily posting cadence and a displayed weekly range, highlighting the need to explain the calculation and keep the plan internally consistent.
- 4:00:10
Defining Content Tone from Professional Context
An AI prompt is being shaped to review a professional profile and the context of a software project, then use that background to define an appropriate content tone. The excerpt ends before the tone or final result is shown.
- 4:05:27
Editing Role-Play Gaming Stories
An explanation of editing GTA role-play content so events feel dramatically connected, using music, sound effects, and filters to increase viewer immersion, with Adobe Premiere Pro mentioned among the tools used.
Got a question about something in this stream?
Ask in the group