LinkedIn Content Strategy & Writing Style
AI Search @ vespa.ai | ex-Hugging Face / Arcee | 4x Kaggle GrandMaster | 158k+ LinkedIn Followers, 100k+ YouTube Subscribers
1 person tracking this creator on Viral Brain
Abhishek Thakur positions himself as a high-utility practitioner who bridges the gap between elite technical mastery and accessible open-source tooling. His content strategy centers on the rapid deployment of "weekend projects" that solve specific developer pain points, such as remote terminal access or automated tabular modeling, while maintaining a vocal stance against over-engineering and AI hype. What makes him notable is his ability to translate his 4x Kaggle Grandmaster expertise into no-code and low-code solutions, effectively democratizing complex machine learning workflows for a broader audience. This unique intersection of high-level engineering and radical simplicity allows him to serve as both a technical authority and a pragmatic filter in the increasingly cluttered AI landscape.
160.0K
11.7K
54
—
6.2
4
2
New project: vespatune 🚀 Vespatune is a web/cli based python application for training tabular models. - Automatic hyperparameter optimization using Optuna - Supports XGBoost, LightGBM, CatBoost, and…
Why are people using AI models (and "agents") to complicate their lives? Most of what you are automating doesn't need automation 😐
Weekend project: WebTerm; fully-featured, light-weight web-based terminal. You can just pip install this and run on your remote machine (workstations or mac mini) to get the full terminal experience…
vespatune: train tabular models without writing a single line of code!!!
6.2 posts/week
Posts / Week
1.3 days
Days Between Posts
2
Total Posts Analyzed
HIGH
Posting Frequency
53.5%
Avg Engagement Rate
STABLE
Performance Trend
45
Avg Length (Words)
MEDIUM
Depth Level
ADVANCED
Expertise Level
0.72/10
Uniqueness Score
YES
Question Usage
0%
Response Rate
Writing style breakdown
<start of post>
New project: CloudDesk ★
CloudDesk is a lightweight web/cli based dashboard for running small “personal infra” tools on any box you have lying around (old laptop, mac mini, workstation, a $5 VPS, whatever). You can spin up a clean UI in minutes, then open it from any browser (ipad, tablet, phone, your work laptop) and manage everything without ssh-ing into five terminals.
Why i built this: i kept re-doing the same setup every weekend. and most of the “AI agent” wrappers i tried made it harder, not easier 😐 so i went the opposite direction: fewer abstractions, more visibility, boring reliability.
App Launcher (run scripts, containers, services)
Built-in Reverse Proxy
Web Terminal (xterm compatible)
File Explorer + Upload/Download
System Monitoring (CPU/RAM/Disk/Network)
Process Viewer + Kill/Restart
Logs Viewer (tail, search, filter)
Secrets Store (local encrypted)
Token Authentication
Multi-user (basic roles)
Dark/Light Themes
Export/Import Config
Works on Linux, macOS, and Windows
What it feels like: open one page, see everything. click to start/stop. copy logs. jump into a terminal. done.
pip install clouddesk
clouddesk start --host 0.0.0.0 --port 8080
If you already have python on the machine, that’s it. no docker required (but supported). and yes, it runs fine on tiny boxes.
Mac mini in a closet as a build runner + artifact host
Remote GPU workstation for training + monitoring
A cheap VPS for cron jobs + backups
Old laptop as a home dashboard
This is not trying to replace kubernetes.
This is not an “agent”.
This is just a simple control panel that you actually want to keep open.
Repo: https://lnkd.in/ew98xGhN
If this is useful, give the repo some love ⭐️
Want a plugin (postgres panel, redis panel, s3 browser, hf spaces deploy)? open an issue or a pull request. i’ll merge good stuff fast :D
<end of post>
Sign in to unlock the full writing analysis
Nail your LinkedIn strategy with ViralBrain.
Analyze and write in Abhishek Thakur's style. Grow your LinkedIn to the next level.