David Tang
banner
davidtang.ai
David Tang
@davidtang.ai
🩺 Clinical AI consultant
🧑‍🤝‍🧑AI community lead in London

🔗 davidtang.ai
Got em?

Cut out the stinky bots withe some strict rules and enforcing certificate handshakes.

These guys were hitting every possible URI, doing vulnerability scans and who knows what else.

Left side all bots ... right side, actual users
June 30, 2025 at 9:18 PM
What's better than vibe coding

Vibing with prism glasses

...

Now I can keep my neck aligned at least

And worry about aligning the codebase later
March 22, 2025 at 9:06 AM
As part of a digital health series for care providers ... I've thrown my hat into the mix with this
@Peerr_io
exclusive

AI x Healthcare: where do the opportunities lie?

What trends are driving in the next few years?

Check out the full video below.
March 11, 2025 at 9:12 AM
- Recording: youtu.be/QDXXYVznb3I?...
- Launch blog post: tinyurl.com/orra-launch-...
- GitHub repo: github.com/orra-dev/orra
- Landing page: orra.dev

Watch this space for more AI agent related calls monthly!
February 25, 2025 at 10:48 AM
A totally 🔥 session with @ezodude.bsky.social
at
AICamp's Agentic Hours last week

Check out the open-source repo that @newnomad.bsky.social
and Ezo are building in the multi-agent space!
February 25, 2025 at 10:48 AM
UX input needed! I think radio buttons are too limiting for selecting more than one type, but "ALL" should check everything?

Is there a better input mechanism for this menu?
February 15, 2025 at 6:27 PM
Saturday build ...

Finally some GUI 🫡
Bit on the fence here, I think "All" should automatically tick the following checkboxes. Then again radio-buttons are a bit too limiting.

UX guys! whats your take

#buildinpublic
February 15, 2025 at 6:25 PM
Was at the RCP museum yesterday for an AI x Healthcare talk.

Got some sick potions in their museum 🍶
February 15, 2025 at 11:47 AM
My booster for today:
A mix of matcha with salted caramel latte.

🍵 perfect companion for vibecoding.

Also 🙏 high 5 to the latest v1.0 bump for the google-genai Python SDK. Loving it and integrating it into my pipeline has never been easier.
February 12, 2025 at 11:59 AM
Keep trucking on, fellow builders.
Escape the valley of despair.
#buildinpublic

Often the initial excitement wears out;

Things get boggy, quicksand everywhere.

Each attempt, I think - gets you further out of the valley.

Keep going, keep trucking on.
February 10, 2025 at 10:59 AM
#buildinpublic

This week was spent hacking together a few tools for LLMs to use:

🔹pdf parser
🔸video-to-text via whisper
🔹good ol' RAG

I'm just happy I don't have to shift through that deng old musty document with over 1000+ variable names now
🫡
February 7, 2025 at 10:13 PM
Hey blue #healthtech !
If you are around please join our panelist on the following topic

26 Feb / Manchester Central
February 7, 2025 at 12:13 PM
Is AI the 𝗺𝗶𝘀𝘀𝗶𝗻𝗴 𝘀𝗹𝗶𝗰𝗲 in Healthcare?

The "Swiss Cheese Model" 🧀 shows how layers of safeguards prevent errors in healthcare.

AI could add a new layer of protection OR create new holes.

🤔Things to consider:

Will it increase cognitive load on staff?
Where does the AI layer fit best?
February 6, 2025 at 2:47 PM
A less than glorious day building.

Spent hunching over the kitchen counter. My build material today: fine steel wool

Keeping out the 🐁

He's winning so far by beating one barrier 🚧

Let's see how he fares with this.
February 4, 2025 at 12:50 PM
"Vibe coding" -- is this the future of code ⁉️

Either code will lose it's moat;

Or in 5 years programming will be something completely unrecognizable to what it was.
February 4, 2025 at 9:42 AM
Deepseek V3
R1 with Qwen
o3-low
o3-high

Spoilt for choice ... What's your main weapon / building tool?
February 2, 2025 at 11:01 AM
Giving myself and the fam ... permission to breath.

I've been enjoying the different perspective the locals have.

Sometimes being "at it" for too long a time, creates some biases that can be both good and bad.

Where is the balance for you?
January 28, 2025 at 1:30 PM
#buildinpublic

Looking back at progress from 2 weeks ago.
Slow but steady; I'm happy with direction.

Remember you are what you eat!
Garbage in, garbage out.

Watch what you feed your models too.
Give them the nutritious stuff.
January 23, 2025 at 2:18 PM
Kicking off two internal initiatives this week.
Haven't been so pumped for Monday!

Meetings on the calendar which I actually look forwards to.
Focused, and interesting agendas. 🥰

slowly and steadily building on the background tho, LFGG!
January 20, 2025 at 9:19 AM
Playing defence today!

Saw this post about a $450k LLM bill after being compromised! That's insane 😦

Secure your endpoints, keys, lock your houses.
Set your alarms.
January 17, 2025 at 11:36 PM
Just wrapped up our 7-Day AI Immersion at AICamp London.

A self-study, virtual, and instructor guided course that was completely free for over 800+ developers in >5 countries.

We touched on SDKs, frameworks, new IDEs from various players including OpenAI, Gemini, Ollama and prompt engineering!
January 16, 2025 at 10:06 AM
Me ... after getting alpha feedback and speaking to user on their workflows.

I'm guessing many will echo at #buildinpublic
January 15, 2025 at 9:08 PM
30 mins to lunch. Can I push out this one feature?

Lets gooo 🚀🚀🚀

#buildinpublic
January 13, 2025 at 11:33 AM
Its that time again ☕ #coffeehour

Waiting for my franken-baby to go live ...
Lets build!

#buildinpublic
January 9, 2025 at 2:34 PM
AI agents are making waves, poised to transform work.

InfoWorld highlights their potential to revolutionize software development, business processes & customer experiences.

The market's projected to grow 45.1% annually, reaching billions by 2030.
January 7, 2025 at 1:33 PM