Why do websites still hijack scrolling? It sucks

Location: Bengaluru, India

Remote: Yes

Willing to relocate: Yes

Technologies: Languages: Python, Java, JavaScript, Bash AI / ML Serving: model serving over HTTP, async inference pipelines, job queues, edge deployment of DL models Infrastructure: AWS, Docker, CI/CD, Bazel, on-prem deployment automation Data: PostgreSQL, RabbitMQ / queue-based messaging Practices: distributed systems design, API design, incident response, engineering process for regulated/clinical software

Résumé/CV: https://www.arunkant.com/resume/

Email: [email protected]

Engineering leader, 10+ years. Spent 7 years at Qure.AI (medical imaging AI) — built the backend serving deep-learning models, then led the migration of three product lines into a unified Bazel monorepo, then ran engineering for the Global Health BU (~20 engineers, scaled chest X-ray AI to 80+ countries and 2,000+ sites). Left in May 2025, currently building two AI side projects (releasedog.com, Vellum). Looking for Director / Head of Engineering / VP roles at AI/ML startups, or founding-engineer at seed stage.

This is pretty well done

I'm working on https://releasedog.com/ - It collects changes from JIRA/slack/github and create a changelog/release notes.

Why does your agent have permission to delete production database?

It was explained in the post

Did you read the article? They did not believe that the token the agent had access to had the ability to delete production data using it.

They don't.

Yep, it should be better and more efficient then sonnet.

Probably coming next

Maybe just ask for confirmation