DEV Community

vadim albarov profile picture

vadim albarov

Health-tech engineer, building privacy tools for clinical text

Joined Joined on 
Your AI Knows the Rule. Your Question Decides Whether You Get the Myth
vadim_albarov profile

Your AI Knows the Rule. Your Question Decides Whether You Get the Myth

#discuss #ai #security #hipaa
Comments 1
9 min read
How Many Lines of Code Do You Ship to Prod a Week? This Week I Shipped One.
vadim_albarov profile

How Many Lines of Code Do You Ship to Prod a Week? This Week I Shipped One.

#ai #aws #claude #devops
3
Comments
7 min read
Windows Copilot Inverted a HIPAA Rule Three Times - and Cited a Real Paper That Proves the Opposite
vadim_albarov profile

Windows Copilot Inverted a HIPAA Rule Three Times - and Cited a Real Paper That Proves the Opposite

#ai #hipaa #security #llm
2
Comments
9 min read
The De-identification Datasets Problem: i2b2/n2c2 Access Is Broken, and What to Do Instead
vadim_albarov profile

The De-identification Datasets Problem: i2b2/n2c2 Access Is Broken, and What to Do Instead

#healthtech #machinelearning #nlp #opensource
2
Comments
11 min read
Is 'HIPAA-Eligible' the Same as 'HIPAA-Compliant'? I Audited AWS's List Against the Security Rule
vadim_albarov profile

Is 'HIPAA-Eligible' the Same as 'HIPAA-Compliant'? I Audited AWS's List Against the Security Rule

#aws #hipaa #security #healthtech
3
Comments
11 min read
"Oops, I Forgot to Tell You That's Dangerous": Claude Code Watched Me Wipe Production Redis - Then Helped Carve It Back Off the Disk"
vadim_albarov profile

"Oops, I Forgot to Tell You That's Dangerous": Claude Code Watched Me Wipe Production Redis - Then Helped Carve It Back Off the Disk"

#claude #redis #postmortem #devops
2
Comments
11 min read
A Design Flaw in Claude Code's Documentation Skill: One Question, 265k–355k Tokens
vadim_albarov profile

A Design Flaw in Claude Code's Documentation Skill: One Question, 265k–355k Tokens

#ai #claude #llm #debugging
1
Comments 2
8 min read
I fine-tuned a 1.7B model on data that doesn't exist - a 17-minute recipe
vadim_albarov profile

I fine-tuned a 1.7B model on data that doesn't exist - a 17-minute recipe

1
Comments
10 min read
A 125M model beat a 14B LLM at de-identifying medical text 40 faster, on CPU
vadim_albarov profile

A 125M model beat a 14B LLM at de-identifying medical text 40 faster, on CPU

#showdev #ai #machinelearning #privacy
1
Comments
12 min read
loading...