[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, agentic, if you, llm 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% [stocks](/list/stocks)  XXXX% [finance](/list/finance)  XXXX%

**Social topic influence**
[ai](/topic/ai) 13.56%, [agentic](/topic/agentic) #80, [if you](/topic/if-you) 6.78%, [llm](/topic/llm) #21, [math](/topic/math) #503, [mcp server](/topic/mcp-server) #37, [$googl](/topic/$googl) #2246, [just a](/topic/just-a) 3.39%, [virtual](/topic/virtual) 1.69%, [dns](/topic/dns) XXXX%

**Top accounts mentioned or mentioned by**
[@rauljuncov](/creator/undefined) [@codingwithroby](/creator/undefined) [@llamaindex](/creator/undefined) [@huggingface](/creator/undefined) [@systemdesignone](/creator/undefined) [@rasbt](/creator/undefined) [@kundan31564313](/creator/undefined) [@pirouneb](/creator/undefined) [@prontoronto](/creator/undefined) [@_nghiaphan1019](/creator/undefined) [@frankleo61](/creator/undefined) [@principal_ade](/creator/undefined) [@nash](/creator/undefined) [@brandgrowthos](/creator/undefined) [@the_longer_game](/creator/undefined) [@ulidabess](/creator/undefined) [@devanshijais](/creator/undefined) [@calvincalcium](/creator/undefined) [@ollydbarcort](/creator/undefined) [@frankhhq](/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

"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 20K followers, 251.9K 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 20K followers, 2770 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 20K followers, 73.7K 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 20K followers, 18.9K 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 20K followers, 11.7K 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 20K followers, 39.6K 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 20K followers, 60.5K 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 20K followers, 2251 engagements


"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 20K followers, 617.5K 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 20K followers, 138.8K 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 20K followers, 4568 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 20K followers, 3730 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 20K followers, XXX 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


"Here's the playlist: Happy Learning"  
[X Link](https://x.com/techNmak/status/1998062961248882861)  2025-12-08T16:11Z 19.9K followers, XXX 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 20K followers, 18K 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 20K followers, 23.2K engagements


"Meta just solved RAG's biggest bottleneck. XX faster decoding. Zero accuracy loss. The problem nobody talks about: When you feed an LLM XX retrieved passages only 5-10 are actually useful. The rest Dead weight. But you're computing attention for ALL of them. The math is brutal: Traditional RAG with 16K context: 100+ seconds to first token XX throughput drop Massive memory waste What REFRAG does: Compresses context chunks into single embeddings. Instead of processing 16384 tokens Process 1024 chunk embeddings. The results: ✓ XXXXX faster time-to-first-token ✓ Zero perplexity loss ✓ XX context"  
[X Link](https://x.com/techNmak/status/1998847976341647776)  2025-12-10T20:11Z 20K followers, 20.8K 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 20K followers, 56.5K 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 20K followers, 126.3K 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 20K followers, 76.9K 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 20K followers, 98K 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 20K followers, 95.4K 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 20K followers, 1930 engagements


"Agentic frontends just hit a new inflection point. 2025 has already been wild for agentic UI. But today felt like a proper engineering milestone. CopilotKit just released version XXXX and it includes a new useAgent() universal React hook that is a game changer for Agentic applications. CopilotKit unifies the entire Agent stack into one framework so you can build "Cursor for X" style apps without implementing each protocol from scratch. What does useAgent() do - Streams all agent events into your UI (messages tool calls status updates) - Synchronizes your agent and app state - Sends user input"  
[X Link](https://x.com/techNmak/status/1999095329392926898)  2025-12-11T12:34Z 20K followers, 3639 engagements


"@systemdesignone Great books :) DDIA is my all time favorite"  
[X Link](https://x.com/techNmak/status/1999105542854451575)  2025-12-11T13:14Z 20K followers, XXX 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 20K followers, 1741 engagements


"The gold-level resources to learn LLMs from scratch. @rasbt has created these masterpieces available on Manning. I can genuinely say this as Im currently going through the video course myself. I will share the book & companion video course link from Manning in the comments"  
[X Link](https://x.com/techNmak/status/1996322844708635028)  2025-12-03T20:57Z 20K followers, 18.3K 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 20K followers, 43.7K 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 20K followers, 69.7K 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 20K followers, 70.9K 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 20K followers, XXX engagements


"@RaulJuncoV Spot on. It turns 'agentic UI' from a complex research project into just another standard dependency. Huge win for velocity"  
[X Link](https://x.com/techNmak/status/1999108357685457006)  2025-12-11T13:25Z 20K followers, XX engagements