[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

# ![@techNmak Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::1818381581897412608.png) @techNmak Tech with Mak

Tech with Mak posts on X about ai, llm, if you, mak the most. They currently have XXXXXX followers and XX posts still getting attention that total XXXXXX engagements in the last XX hours.

### Engagements: XXXXXX [#](/creator/twitter::1818381581897412608/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1818381581897412608/c:line/m:interactions.svg)

- X Week XXXXXXX -XX%
- X Month XXXXXXXXX +159%
- X Months XXXXXXXXX +117%
- X Year XXXXXXXXX +40,950%

### Mentions: XX [#](/creator/twitter::1818381581897412608/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1818381581897412608/c:line/m:posts_active.svg)

- X Week XX -XX%
- X Month XXX +169%
- X Months XXX +103%
- X Year XXX +1,863%

### Followers: XXXXXX [#](/creator/twitter::1818381581897412608/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1818381581897412608/c:line/m:followers.svg)

- X Week XXXXXX +5.10%
- X Month XXXXXX +36%
- X Months XXXXXX +133%
- X Year XXXXXX +3,608%

### CreatorRank: XXXXXXX [#](/creator/twitter::1818381581897412608/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1818381581897412608/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[technology brands](/list/technology-brands)  XXXX% [finance](/list/finance)  XXXX% [stocks](/list/stocks)  XXXX% [social networks](/list/social-networks)  XXXX%

**Social topic influence**
[ai](/topic/ai) 13.64%, [llm](/topic/llm) #156, [if you](/topic/if-you) 6.06%, [mak](/topic/mak) #90, [math](/topic/math) #1925, [scratch](/topic/scratch) #174, [events](/topic/events) 3.03%, [all in](/topic/all-in) 3.03%, [agentic](/topic/agentic) #370, [ai agent](/topic/ai-agent) #9

**Top accounts mentioned or mentioned by**
[@huggingface](/creator/undefined) [@copilotkit](/creator/undefined) [@codingwithroby](/creator/undefined) [@dhanushnehru](/creator/undefined) [@llamaindex](/creator/undefined) [@pirouneb](/creator/undefined) [@_nghiaphan1019](/creator/undefined) [@prontoronto](/creator/undefined) [@frankleo61](/creator/undefined) [@principal_ade](/creator/undefined) [@brandgrowthos](/creator/undefined) [@nash](/creator/undefined) [@ulidabess](/creator/undefined) [@the_longer_game](/creator/undefined) [@calvincalcium](/creator/undefined) [@devanshijais](/creator/undefined) [@ollydbarcort](/creator/undefined) [@frankhhq](/creator/undefined) [@syedaadamahmad](/creator/undefined) [@optimumdwc](/creator/undefined)

**Top assets mentioned**
[Alphabet Inc Class A (GOOGL)](/topic/$googl) [Microsoft Corp. (MSFT)](/topic/microsoft)
### Top Social Posts
Top posts by engagements in the last XX hours

"There are X career paths in AI right now: The API Caller: Knows how to use an API. (Low leverage first to be automated $150k salary). The Architect: Knows how to build the API. (High leverage builds the tools $500k+ salary). Bootcamps train you to be an API Caller. This free 17-video Stanford course trains you to be an Architect. It's CS336: Language Modeling from Scratch. The syllabus is pure signal no noise: ➡ Data Collection & Curation (Lec 13-14) ➡ Building Transformers & MoE (Lec 3-4) ➡ Making it fast (Lec 5-8: GPUs Kernels Parallelism) ➡ Making it work (Lec 10: Inference) ➡ Making it"  
[X Link](https://x.com/techNmak/status/1990802817305477448)  2025-11-18T15:22Z 19.9K followers, 617K engagements


"These GitHub repos have leveled up my AI skills more than any bootcamp course or conference. If you're serious about AI these are the ones worth bookmarking 👇"  
[X Link](https://x.com/techNmak/status/1994416682383905206)  2025-11-28T14:42Z 19.9K followers, 73.6K engagements


"I still cant believe this is free. Most bootcamps are charging $3000 to teach you outdated material. Meanwhile @huggingface is giving away the state-of-the-art curriculum for $X. Agents ✅ Robotics ✅ The new MCP standard ✅ Check this. Bookmark.👇"  
[X Link](https://x.com/techNmak/status/1994793715332956501)  2025-11-29T15:40Z 19.9K followers, 126.1K engagements


"If you are preparing for your System Design Interview these resources will be very helpful for you. 👇 Read it. Bookmark it"  
[X Link](https://x.com/techNmak/status/1995009289535271422)  2025-11-30T05:57Z 19.9K followers, 76.8K engagements


"This GitHub README is better than most $XXX AI courses. Read it. Bookmark it.👇"  
[X Link](https://x.com/techNmak/status/1995187164100096477)  2025-11-30T17:44Z 19.9K followers, 97.8K engagements


"LLM Prompting Techniques Read. Learn. Bookmark. Prompting isnt just asking the AI a question. Its a deliberate engineered input design process and a critical skill when working with Large Language Models (LLMs). Let's breakdown the prompting techniques. ✅ X. Core Prompting Techniques Zero-shot - No examples provided. Just the task. One-shot - One example shown before the task. Few-shot - A handful of examples used to teach patterns. 🧠 X. Reasoning-Enhancing Techniques Chain-of-Thought (CoT) - Encourage step-by-step reasoning. Self-Consistency - Sample multiple CoTs; choose the best."  
[X Link](https://x.com/techNmak/status/1995726428177137924)  2025-12-02T05:27Z 19.9K followers, 43.6K engagements


"🚀 RAG has evolved far beyond its original form. When people hear Retrieval-Augmented Generation (RAG) they often think of the classic setup: retrieve documents feed into LLM generate an answer. But in practice RAG has branched into many specialized patterns each designed to solve different challenges around accuracy latency compliance and context. Here are some of the most important categories: ➤ Standard RAG - the original retrieval + generation (RAG-Sequence RAG-Token). ➤ Graph RAG - connects LLMs with knowledge graphs for structured reasoning. ➤ Memory-Augmented RAG - external memory for"  
[X Link](https://x.com/techNmak/status/1997174041690423462)  2025-12-06T05:19Z 19.9K followers, 23.1K engagements


"I have one interview question I use to find real ML engineers: "Explain Backpropagation. No not the concept. The math. From scratch." X out of XX candidates can't. They can use a library. They can't build one. The 1/10 who can They've all built the foundation. This 26-video playlist is that foundation. For free. While everyone else is chasing the newest "AI agent" or prompt hack they're building on a foundation of sand. This free course from Professor Bryce is the foundation. It's a full university-level curriculum on the math that actually makes AI work. The syllabus is pure signal no noise:"  
[X Link](https://x.com/techNmak/status/1992005231031484884)  2025-11-21T23:00Z 19.7K followers, 98.4K engagements


"My doctor told me to reduce stress. I replaced Apache with Nginx. 😉 Nginx (pronounced "engine-x") - - Web Server - Reverse Proxy - Load Balancer NGINX is a reverse proxy load balancer meaning it manages connections between clients and backend servers. It is free & open-source. It's renowned for its efficiency stability and ability to handle massive loads concurrently.  - At its heart Nginx is an event-driven web server. It doesn't dedicate a thread to each incoming connection (like traditional models). Instead it relies on a single (or a few) worker processes to manage multiple connections"  
[X Link](https://x.com/techNmak/status/1993193797426331993)  2025-11-25T05:43Z 19.8K followers, 28.2K engagements


"Every LLM app needs evaluation. Most teams build it from scratch. Opik is the open-source platform that does it for you. Tracing. Evaluation. Monitoring. Agent optimization. Guardrails. Integrates with everything: OpenAI LangChain LlamaIndex Anthropic 50+ more. Stop reinventing the wheel. (GitHub in comments)"  
[X Link](https://x.com/techNmak/status/1994123635305640246)  2025-11-27T19:18Z 19.8K followers, 14.2K engagements


"I have created this illustration to help you visualize the Docker Workflow 👇 Let's understand the terms using analogy - 👉 Dockerfile - Think of a Dockerfile as a recipe or a set of instructions. You start by creating a Dockerfile that lists all the ingredients (software and configurations) needed for your application. 👉 Docker Image - Using the Dockerfile as your recipe you "cook" or "build" a Docker Image. This image is like a frozen snapshot of your application containing everything it needs to run. 👉 Docker Container - Once you have your Docker Image you can "serve" it by creating a"  
[X Link](https://x.com/techNmak/status/1996128460994171289)  2025-12-03T08:04Z 19.7K followers, 7211 engagements


"Design Patterns XXX Check bonus in the end of the tweet   X. Reusability proven solutions that work across projects X. Flexibility encourage loose coupling and easier maintenance X. Communication provide a shared vocabulary for discussing design issues and solutions X. Experience encapsulate the experience of seasoned developers helping others avoid common pitfalls and design better software 📌  X. Abstract Factory creates families of related objects through a common interface X. Builder separates complex object construction from representation X. Factory Method defines an interface for"  
[X Link](https://x.com/techNmak/status/1997929852247376075)  2025-12-08T07:22Z 19.8K followers, 10.6K engagements


"LLM Generation Parameters These are the primary controls used to influence the output of a Large Language Model. 1./ Temperature Controls the randomness of the output. It is applied to the probability distribution of the next possible tokens. Low Temperature (e.g. 0.2): Makes the output more deterministic and focused. The model will almost always select the most probable next token. (ideal for factual tasks like summarization code generation and direct Q&A) High Temperature (e.g. 1.0): Makes the output more random and creative. The model is more likely to select less probable tokens leading"  
[X Link](https://x.com/techNmak/status/1997985083639345297)  2025-12-08T11:02Z 19.8K followers, 8293 engagements


"Bonus : Let me tell you about a session-recording tool for developers that captures UI interactions backend logs traces and network events all in one timeline. Super helpful for reproducing intermittent bugs. Give it a try while it has free (forever) plan"  
[X Link](https://x.com/techNmak/status/1998062964323258464)  2025-12-08T16:11Z 19.7K followers, XXX engagements


"If you work with databases especially as a software developer it's important to understand the difference between partitioning and sharding. 📌 = the process of splitting a large table or index into smaller more manageable pieces called partitions. But remember These partitions reside within the same database instance sharing the same resources and management system. They cannot span multiple databases. But they can be distributed across multiple storage devices within the same server. The database system parses the incoming query and based on the partition key values it determines which"  
[X Link](https://x.com/techNmak/status/1976941749017759757)  2025-10-11T09:23Z 19.9K followers, 56.4K engagements


"Why "Delete" doesn't actually Delete (the tombstone trap) In Log-Structured Merge (LSM) databases like Cassandra ScyllaDB or RocksDB files are immutable. Once written they cannot be modified. So how do you delete a record You write a new one. 1./ The Tombstone To delete User123 the database writes a new record with a special marker: Key: User123 Value: TOMBSTONE A Tombstone is effectively a note that says: "This key is dead as of 10:05 AM." X. /The Read Path When you query data the database reads both the old record and the new marker: User123: "Alice" (Timestamp: 10:00) User123: TOMBSTONE"  
[X Link](https://x.com/techNmak/status/1991479627869741428)  2025-11-20T12:11Z 19.9K followers, 1741 engagements


"These engineering blogs have leveled up my tech skills more than any bootcamp course or conference. Here are the ones worth bookmarking:"  
[X Link](https://x.com/techNmak/status/1992690826049474661)  2025-11-23T20:24Z 19.9K followers, 251.7K engagements


"This free CUDA course is worth more than most CS degrees. XX hours that separate library users from GPU engineers. I watched senior devs struggle with concepts taught in hour X. What makes it different: No hand-waving. No "just use this library." You build an MLP trainer FOUR times: PyTorch (the easy way) NumPy (getting harder) C (now we're cooking) CUDA (chef's kiss) Same model. Same dataset. Four implementations. By the end you understand WHY PyTorch is fast. The curriculum nobody else teaches: ➡GPU architecture (not just "it's parallel") ➡Writing kernels that don't suck ➡Profiling at"  
[X Link](https://x.com/techNmak/status/1993985225014043046)  2025-11-27T10:08Z 19.9K followers, 138.7K engagements


"System design is about creating applications that can handle real-world demands.  📌 DNS - Domain Name System (resolvers nameservers records) Load Balancers - Hardware software Layer X Layer X CDNs - Content Delivery Networks (caching edge servers) Proxies - Forward reverse transparent anonymous VPNs - Virtual Private Networks (tunneling protocols) Firewalls - Packet filtering stateful inspection NAT - Network Address Translation Gateways - Connect different networks Routers - Direct traffic between networks 📌 Databases - SQL NoSQL (key-value document columnar graph) NewSQL Object Storage -"  
[X Link](https://x.com/techNmak/status/1994555410674831820)  2025-11-28T23:54Z 19.9K followers, 18.8K engagements


"An SLM is a Small Language Model usually under 10B parameters. Runs on consumer hardware. Low latency. Low cost. High control. And according to NVIDIA Research these not giant 100B+ LLMs are what will actually power agentic systems. Heres the distilled insight 👇 1./ We built agents wrong. We assumed an Agent needs a giant model that "knows everything." But real agents do boring repetitive work: parse a command call a tool extract JSON run again Using a frontier LLM for this loop is like hiring a PhD to press elevator buttons. It works but its economically insane. 2./ Why SLMs win (the 3"  
[X Link](https://x.com/techNmak/status/1995262353840218206)  2025-11-30T22:43Z 19.8K followers, 11.6K engagements


"Every software developer should know the difference. Concurrency and Parallelism are not the same. Follow the 🧵"  
[X Link](https://x.com/techNmak/status/1995483885082210406)  2025-12-01T13:23Z 19.8K followers, 22.2K engagements


"LLM = Smart but works only with what it remembers from training. RAG = Smart + can look things up to be more informed and up-to-date. Agent = Smart + can decide what to do next to achieve a goal using tools memory or plans. ------ Bonus: Let me tell you about a session-recording tool for developers that captures UI interactions backend logs traces and network events all in one timeline. Super helpful for reproducing intermittent bugs. (Give it a try for free)"  
[X Link](https://x.com/techNmak/status/1995562342000853149)  2025-12-01T18:35Z 19.9K followers, 18K engagements


"I have one interview question I use to find real ML engineers: "Explain Backpropagation. No not the concept. The math. From scratch." X out of XX candidates can't. They can use a library. They can't build one. The 1/10 who can They've all built the foundation. This 26-video playlist is that foundation. For free. While everyone else is chasing the newest "AI agent" or prompt hack they're building on a foundation of sand. This free course from Professor Bryce is the foundation. It's a full university-level curriculum on the math that actually makes AI work. The syllabus is pure signal no noise:"  
[X Link](https://x.com/techNmak/status/1995872587403198667)  2025-12-02T15:08Z 19.9K followers, 39.5K engagements


"First Google then Microsoft and now AWS It seems like every week one of the tech giants is integrating with the same protocol. If you havent been following - Im talking about AG-UI AG-UI (the Agent-User Interaction protocol) connects any agentic backend to the frontend. Itis a general-purpose bi-directional connection between a user-facing application and any agentic backend. AG-UI has first party integrations & partnerships with Googles ADK Microsofts Agent Framework AWS Strands LangGraph CrewAI PydanticAI Mastra LlamaIndex and more. And CopilotKit the company behind AG-UI provides"  
[X Link](https://x.com/techNmak/status/1996426588553089506)  2025-12-04T03:49Z 19.9K followers, 95.3K engagements


"Make the most of your weekend. Don't sleep on this. Stanford's Autumn 2025 Transformers & LLMs course. X lectures. Free. While others scroll you could understand how Flash Attention achieves 3x speedup how LoRA cuts fine-tuning costs by XX% and how MoE makes models efficient. ➕ What's covered: ➡ Lecture 1: Transformer Fundamentals Tokenization and word representation Self-attention mechanism explained Complete transformer architecture Detailed implementation example ➡ Lecture 2: Advanced Transformer Techniques Position embeddings (RoPE ALiBi T5 bias) Layer normalization and sparse attention"  
[X Link](https://x.com/techNmak/status/1997293275569631644)  2025-12-06T13:13Z 19.9K followers, 60.3K engagements


"Stop trying to learn complex Math and AI concepts from static PDF files. 🛑 Interactive visualization is the cheat code for deep understanding. Here are X incredible free resources to master Linear Algebra Probability and Deep Learning visually. 🧵👇"  
[X Link](https://x.com/techNmak/status/1997657844468899904)  2025-12-07T13:21Z 19.9K followers, 69.4K engagements


"Here's the playlist: Happy Learning"  
[X Link](https://x.com/techNmak/status/1998062961248882861)  2025-12-08T16:11Z 19.9K followers, XXX engagements


"Last month Google dropped something interesting: five AI Agent papers released across five consecutive days one per day each digging into a different part of how agents should be built evaluated secured and deployed. No big splash just a steady rollout of more than XXX pages of highly technical material. Heres the distilled version of what those five drops covered: Read. Learn. Bookmark"  
[X Link](https://x.com/techNmak/status/1998264904563007889)  2025-12-09T05:34Z 19.9K followers, 68.7K engagements


"1./ The first paper reframed what an agent actually is. It walked through how agent capabilities progress over time and why most current agents struggle the moment they leave demo environments. 🔗"  
[X Link](https://x.com/techNmak/status/1998264909973639399)  2025-12-09T05:34Z 19.9K followers, 3549 engagements


"2./ The second release focused on tooling and MCP. One standout point: an MCP server can add tools without asking you. Great for capability expansion risky if youre not watching the boundaries. 🔗"  
[X Link](https://x.com/techNmak/status/1998264914658672745)  2025-12-09T05:34Z 19.9K followers, 2634 engagements


"3./ The third paper dug into memory - real memory. Not transcripts or logs but structured long-term memory that shapes future reasoning and behavior. 🔗"  
[X Link](https://x.com/techNmak/status/1998264919897395482)  2025-12-09T05:34Z 19.9K followers, 2138 engagements


"4./ The fourth paper tackled agent quality. Their position was clear: you can test software but you cant test judgment the same way. So their framework evaluates how an agent reasons not just what it outputs. 🔗"  
[X Link](https://x.com/techNmak/status/1998264924397838412)  2025-12-09T05:34Z 19.9K followers, 1830 engagements


"The biggest bottleneck in RAG isnt your vector DB or your embedding model. Its your Ingestion Pipeline. 📉 @llama_index just solved the problem. If you feed a single PDF containing XX different invoices into your retrieval system you are polluting your context window. Fixed-size chunking is blind. LlamaSplit is semantic. It splits "Frankenstein" documents into discrete logical units using Natural Language instructions. Input: blob .pdf Instruction: "Split into individual tax forms." Output: form_1 form_2 form_3 This is how you build reliable Document Agents. Check comments for the Getting"  
[X Link](https://x.com/techNmak/status/1998697176776716331)  2025-12-10T10:11Z 19.9K followers, 3487 engagements


"The killer feature is their 'Local Rails' infrastructure. Traditional banks use SWIFT (slow expensive). Airwallex spent XX years connecting directly to local banking systems across the globe. Result XX% of payments bypass SWIFT Money moves instantly You save 4-6% on FX fees globally"  
[X Link](https://x.com/techNmak/status/1998765911201951892)  2025-12-10T14:45Z 19.9K followers, XXX engagements


"For developers the integration experience is next-level. Using the Airwallex MCP server you can use AI to 'text-to-code' your billing setup. Watch this = A complete pricing page integration generated with just a few prompts. No manual boilerplate needed"  
[X Link](https://x.com/techNmak/status/1998765916168044549)  2025-12-10T14:45Z 19.9K followers, XXX engagements