Technical creator. Involved in everything AI, data, and tech
Based in Nigeria
- Tweets 1,415
- Following 199
- Followers 79
- Likes 14,014
Your favorite tech creator retweeted
An AI agent can find the right URL and still give you the wrong answer.
I ran through a tutorial on building a web-aware agent with the OpenAI Agents SDK, and the product-price example shows why.
Search finds the product page, but the current price may be rendered client-side.
1
1
2
111
Your favorite tech creator retweeted
So the agent knows where the product is, but not what it currently costs.
The pattern I liked was giving it two tools:
Search → discover the page
Fetch → retrieve the current page content
How are your agents handling this gap?
zenrows.com/blog/web-aware-a…
1
2
10
Your favorite tech creator retweeted
“Managed Postgres” can mean very different things.
At minimum, I’d expect patching, scaling, failover + backups to leave the team’s plate.
Databricks’ Lakebase adds serverless scaling, PITR, branching, PgBouncer + pgvector.
databricks.com/blog/managed-…
1
2
41
I built a RAG pipeline that retrieves live web data before generating answers.
The goal was simple.
I wanted a pipeline that could work with information that changes, instead of relying on a static collection of documents.
Here's how I approached it. 🧵
1
3
3
63
Microsoft replaced Azure Playwright Testing with Playwright Workspaces in March 2026. Your tests still run. What you lost is the reporting layer. There's no cross-run history, stability trends, or consolidated failure view.
1
1
3
385
Hackmamba is hiring a Technical Video Content Creator.
Full-time. Remote in Europe. We make videos for B2B SaaS and developer tools, you'd be the face of that work. Best if you have engineering background.
Apply 👇
hackmamba.io/career/technica…
4
815
Most AI tools generate a finished UI and leave you with a single block of code that is hard to test, hard to refactor, and frustrating to move into a real codebase.This walkthrough builds a full chat app where the AI layer, backend, and UI are kept separate by design from scratch
1
2
39
Every AI coding tool rushes to show you a working UI. Bit Cloud stops at architecture first. It defines the component boundaries, state ownership, and data flow before writing a single line of code .
1
2
63
Lovable is great for first release. The problems show up the moment you try to add a second feature.
Tightly connected files, no test coverage, unclear versioning. The code runs. The system is not built to last.
2
1
5
98
A RAG system passed every eval. Faithfulness scored 0.89. The dashboard was green. Two weeks into production, 35% of users were getting wrong answers.
The tests used formal queries. Production queries were casual, so the metrics caught hallucinations but missed intent.
#rag
1
2
48
Every AI code editor in 2026 promises to write your entire app and eliminate debugging forever.
A small number actually help you ship cleaner code faster. The rest are marketing.
1
3
38
I'm definitely bringing my hardest problem
We're excited to announce VectorAI DB, the first vector database purpose-built for high-performance, reliable AI at the edge.
RAG isn't dead. It just can't run in the environments that need it most.
In manufacturing, 46% of AI pilots never leave the OT network. Healthcare, defense, and financial services are no different.
VectorAI DB runs RAG pipelines, semantic search, and real-time AI agents on-premises, at the edge, or air-gapped.
Before today, 1,000+ devs across three hackathons had already built on it. A maritime AI system. An on-device AI therapist. Cardiac imaging on a closed hospital cluster.
And these are just a handful of what VectorAI DB makes possible.
VectorAI DB is NOW OPEN to the public. Come throw your hardest problem at it.
Download link in the comments.
2
16
You're scanning a folder with 300 log files, 12 CSV exports, and a bunch of timestamped backups.
Writing a loop to filter through all of that sounds miserable. Hardcoding filenames is worse. They change constantly.
Python's glob module handles this differently.
2
1
4
351
