@TechWithTimm
Account based inViet Nam
About this account
- Account based in
- Viet Nam
- Connected via
- Viet Nam App Store
Account-level information from X, not a live location or the device used for a specific post.
Software Engineer - YouTuber - AI Enthusiast
Dubai, UAE
Joined August 2018
- Tweets1.5K
- Following78
- Followers16.3K
- Likes1K
If you want to run AI locally with LM Studio, watch this
If you want to run AI locally, your hardware basically decides what models you can use.
Let me show you how to check your VRAM or unified memory, pick the right model, and configure it so it actually runs well.
From there, we’ll connect LM Studio to VS Code and use the local model from another application.
Timestamps:
00:00 | LM Studio
00:23 | What is LM Studio?
00:44 | LM Studio Download
01:16 | Downloading a Model
05:11 | Live Masterclass
06:09 | Loading Models
08:04 | Loading Multiple Models
09:00 | Using Models in LM Studio
11:24 | Loading Large Models
13:50 | Using Models from VSCode
16:32 | Calling the API from Code
Only about 3% of Python job listings ask for Python alone.
Most roles expect you to write maintainable code, work with real data, build APIs, test and debug software, and increasingly build with AI.
Here are the 5 Python skills you need to focus on to get hired in 2026.
00:00 | Overview
00:47 | What Changed?
01:37 | Skill #1
03:15 | Skill #2
04:48 | How to Learn It
06:18 | Skill #3
07:36 | Skill #4
08:56 | Skill #5
10:43 | How to Get Hired
A creator friend makes up to $150,000 on a single YouTube video.
He only publishes twice a month. Each upload is a massive 2 to 4 hour coding build stacked with 3 or 4 sponsor integrations.
It sounds incredible, but it is psychologically brutal to run.
More importantly, volume eliminates variance.
If two uploads completely flop, you still have two others carrying the month. You stop living and dying by whether the algorithm favored you this week.
If you're not sure which AI agent framework to use, watch this
Here’s how to build the same AI coding agent with CrewAI, the OpenAI Agents SDK, and LangGraph.
We’ll go from the simplest setup to the most flexible one, covering tools, multi-agent handoffs, guardrails, state, review loops, and checkpoints along the way.
Then I’ll show you where each framework actually makes sense.
Timestamps:
00:00 | Overview
02:00 | AI Agent Frameworks
03:25 | Way #1 — CrewAI
12:25 | Way #2 — OpenAI Agents SDK
19:39 | Way #3 — LangGraph
30:25 | Which Framework to Use?
If you want to run AI locally, your hardware basically decides what models you can use.
Let me show you how to check your VRAM or unified memory, pick the right model, and configure it so it actually runs well.
From there, we’ll connect LM Studio to VS Code and use the local model from another application.
Timestamps:
00:00 | LM Studio
00:23 | What is LM Studio?
00:44 | LM Studio Download
01:16 | Downloading a Model
05:11 | Live Masterclass
06:09 | Loading Models
08:04 | Loading Multiple Models
09:00 | Using Models in LM Studio
11:24 | Loading Large Models
13:50 | Using Models from VSCode
16:32 | Calling the API from Code
If you want to understand how MCP actually works in production, watch this
An MCP server is basically a standardized way to expose tools to an AI client.
In this video, I’ll show you how the client discovers those tools, how tool calls get passed back and forth, and how to build the whole thing in Python.
Then we’ll move from local stdio to HTTP and add authentication, scopes, and per-user access.
Timestamps:
00:00 | Overview
00:15 | What is an MCP Server
03:08 | Local vs Remote
04:13 | Access Control
09:18 | MCP Protocol
11:52 | Building the Server
20:20 | Adding Authentication
Your portfolio only has about 20 seconds to give a recruiter a reason to keep looking.
They are going to skim, not dig through everything.
So if you’re applying right now, watch this before you send out another application.
