Alex Xu shares system design, AI agents, and engineering fundamentals via visual explainers at high frequency.
Save this audit as markdown
Drop it into Notion, Obsidian, or your content plan to use as a working reference.
Activity
Posting style
How they open posts
The first 220 characters of a LinkedIn post are what shows in the feed — these are the recurring shapes they use to get the click.
Comparative Framework
Opens by naming two or three concepts side-by-side to set up a comparison or trade-off analysis.
“REST vs GraphQL vs gRPC”
“Single Agent vs. Multi-Agent Architecture”
“RAG vs Graph RAG vs Agentic RAG”
Problem Statement
Leads with a pain point or challenge that practitioners face, grounding the explainer in real problems.
Definition or Concept Hook
Opens with a straightforward explanation of what something is, often in one sentence or phrase.
“An AI agent can be thought of as a simple While-loop.”
“Latency vs Throughput vs Bandwidth”
Industry Insight or Case Study
Begins by referencing a company's approach or a concrete deployment, then drilling into the mechanics.
Trend or Timely Reference
Opens by highlighting what's relevant right now or what's emerging in the space.
“System design fundamentals have become more important than ever in the age of AI. Let's review some of the most popular Kafka use cases today.”
“NVIDIA's Open Models in 1 Diagram”
Content pillars
AI Agents & LLM Architecture
3600%In-depth explorations of AI agent design, LLM evaluation, memory systems, and production deployment patterns. Covers agent loops, multi-agent vs single-agent systems, RAG variants, and real-world case studies from Salesforce, OpenAI, and Microsoft.
Core System Design & Infrastructure
2600%Foundational technical concepts including databases, caching, APIs, networking protocols, containerization, and performance metrics. Teaches latency vs throughput, proxy patterns, Docker internals, and data structure fundamentals.
AI Model Landscape & Optimization
1900%Surveys of frontier AI models, cost optimization techniques, inference strategies, and hardware comparisons. Covers SLMs vs LLMs, token spend reduction, model routing, and emerging model architectures.
Who they comment on
These are the creators they show up most consistently in the comments of. Showing up in the same conversations puts you in front of the same audience.
What you can take from this
Use side-by-side comparisons (e.g., "A vs B vs C") as a foundational hook; 52% of top-performing posts employed this structure, driving 1,200+ avg reactions.
Lead with real production pain points before explaining the solution (e.g., "Why X Fails...", "Token Spend Out of Control..."); these frames consistently achieved 70+ comments.
Pair textual explainers with single clean infographics; 85% of posts used this format and generated higher engagement than text-only (avg 1,100 reactions vs 306).
Reference real-world case studies from recognizable companies (OpenAI, Salesforce, Microsoft) early in the hook to establish credibility and draw the audience in.
Maintain a 4.5-post-per-week cadence across Mon-Fri with no weekend posting, creating predictable mid-week peaks and reinforcing audience habit formation.