Tu Ba Khuym
@TuBaKhuym
Followers
9
Following
113
Media
7
Statuses
220
Xin chào mình là Tù Bà Khuỳm! Một Backend Engineer trong mảng Mobile Games.
Hà Nội
Joined August 2024
I swear if Google Chrome or safari just added vertical tabs that exactly copied the functionality from Arc everyone would stop caring about every other browser
200
71
2K
OpenCode now supports ACP thanks to @josephschmitt (v0.15.10 and up) This is me using it in zed
18
15
201
To: founders@ Subject: SAFEs You can now send, sign, and track SAFEs in a few clicks, right from the @atlas dashboard. Sincerely, The Stripe team
85
61
1K
There's a new subagent in Amp called the Librarian. Its job is to locate relevant context in all the libraries you might be using in open source or your private dependencies. Amp's accessible context now extends to the entire universe of code: https://t.co/6zgXMo8EZ2
8
13
145
Most asked questions tomorrow: Can we fix it? Can we change to something else, switch from AWS? How can we mitigate the user impact?
0
0
1
Guys I need you answers for this question on the code-red report related to AWS outage. How can we prevent it from happening again?
0
0
0
🧠 LangGraph × cognee Integration cognee brings persistent memory to LangGraph agents, letting AI applications maintain context across sessions while seamlessly working with existing LangGraph features. Check out how to add memory to your agents 🔗 https://t.co/Om6FZE2gRR
7
35
238
Day 7: I am so obsessed with Docling, must try it first. So I deployed the docling-serve image to the A6000 via Brev @NVIDIAAIDev, then ran all the PDFs to perform the conversion. But it failed when I tried to ingest about 10 DoclingDocuments into Cognee. The EC2 instance just
Days 5-6: The crawled documents were in DOCX format and contained tables split across pages and Vietnamese special characters, causing issues when ingested directly into @cognee_. Converting to PDFs resolved the issue. I asked Droid @FactoryAI to convert all DOCX files in
0
0
0
uv is the best thing that has happened to Python devs! And you won't find a better cheatsheet than this:
25
156
1K
macOS Sandboxes now in Early Access on @daytonaio 💻
9
5
83
Days 5-6: The crawled documents were in DOCX format and contained tables split across pages and Vietnamese special characters, causing issues when ingested directly into @cognee_. Converting to PDFs resolved the issue. I asked Droid @FactoryAI to convert all DOCX files in
Day 4: Feed @crawl4ai docs to @FactoryAI also with targeted sites. Droid nailed it, extracted the admin procedure documents after 30 minutes on autopilots. Next step would be cleaning up the documents.
0
0
3
Better T Stack just got some love from Vietnam! 🙌 Watch him build a fully working Next.js + Convex + Clerk app in minutes 😍
1
2
20
Day 4: Feed @crawl4ai docs to @FactoryAI also with targeted sites. Droid nailed it, extracted the admin procedure documents after 30 minutes on autopilots. Next step would be cleaning up the documents.
Day 2 + 3: Bootstrap a new project with better-t-stack from @amanvarshney01, deployed @cognee_ to an ec2 instance and hook it to the stack. Gonna play around first 🙏
1
1
7
Day 2 + 3: Bootstrap a new project with better-t-stack from @amanvarshney01, deployed @cognee_ to an ec2 instance and hook it to the stack. Gonna play around first 🙏
Day 1: Admin procedures are a painful process, especially for someone like me who has no prior knowledge of them. It takes time, effort, and headaches every single time. That’s why I want to create a simple application for people who have lost their way when it comes to
3
1
4
New in the Perplexity Search API — you can now filter searches by specific domains. Query only trusted sources to get focused, verifiable results.
10
11
142
My first tutorial on creative task optimization with DSPy and GEPA is online! Not everything went as expected, but it’s a solid start toward better results. https://t.co/3UaaUCCcCB
Finished my first tutorial on improving the quality of smaller LLMs for creative tasks using a teacher + GEPA. I'll publish it on Monday. What style of signature do you prefer? I usually go with classes, but for this tutorial I decided to use inline
5
38
231