alwinaugustin

I'm searching for senior engineering roles across Europe and need visa sponsorship (my current work permit expires soon). What surprises me is that in 2026, most major job platforms still don't have a way to filter for this upfront.

The problem:

On LinkedIn, Indeed, Glassdoor, and most company career pages, you only discover visa policies after spending time on an application. The typical flow:

1. Find an interesting role matching your skills 2. Spend time customizing resume/cover letter 3. Fill out lengthy application form 4. Hit the question: "Are you authorized to work in [country]?" - Sometimes in the final page

5. Select "No, I will require sponsorship" (Often) Immediate rejection or find out later the company doesn't sponsor

I estimate I've wasted time on applications where sponsorship wasn't available—time I could have spent on actual opportunities.

What works better:

Platforms like Wellfound (AngelList) have a sponsorship checkbox

Relocate.me only lists sponsorship jobs

Honeypot and Otta show visa status upfront for EU roles

But their coverage is limited compared to LinkedIn's scale.

Why this matters:

For candidates: Wastes time on non-opportunities, forces spray-and-pray approach

For companies: Miss qualified international talent, receive applications from candidates they can't hire

For the industry: We say tech talent is global, but our tools make global matching inefficient

What would help:

A simple standardized filter on major platforms: "Visa sponsorship available" (yes/no/case-by-case). Show it before candidates invest time applying.

My questions:

Has anyone found better ways to search for visa sponsorship roles?

Hiring managers: What prevents making this more visible upfront?

Is anyone building (or interested in building) tools to solve this?

This seems like a clear win for matching efficiency. Am I missing something, or is this just an overlooked UX problem?

TLDR: Most job platforms still lack visa sponsorship filters in 2026. You discover policies only after lengthy applications, wasting time for both sides. Niche platforms do this better but have limited coverage. Would love to hear solutions or why this problem persists.

Remotedays helps companies maintain compliance with EU cross-border remote work regulations. When employees work remotely across borders (France, Belgium, Germany, Luxembourg), exceeding specific thresholds triggers social security and tax liability shifts for both employees and employers.

Thresholds:

France/Belgium: 34 days/year Germany: 183 days/year

Most companies track this manually or not at all, creating audit and penalty risk.

Key Features

For Employees:

One-click daily declarations via automated email prompts with real-time threshold tracking

For HR/Compliance:

Real-time compliance dashboard with alerts across the entire workforce

Complete audit trail for regulatory inspections

Links

Live: https://remotedays.app

Demo: https://demo.remotedays.app

Built with security and GDPR compliance as core requirements. Currently seeking feedback and open to customization or developing similar compliance solutions for specific organizational needs.

Questions and feedback welcome.

There's a lot of talk about "AI code generation," but if AI were truly intelligent—able to understand and solve problems autonomously—would we still need programming at all?

Programming is how we translate understanding into instructions. If AI had intelligence, wouldn’t it just solve problems directly, without needing prompts or code?

Current tools still depend on human guidance and structured inputs. So is this really “intelligence,” or are we anthropomorphizing what’s essentially pattern-based automation?

Should we start calling it what it is—statistical code synthesis—rather than framing it as "intelligent" code generation?

I am attempting to automate the PR review process using locally running LLMs. I provide the LLM with the diff and the full file for context. However, the response usually consists of code explanations and general suggestions (e.g., logging, security) rather than an in-depth review covering complexity, alternative logic, or better approaches. Could this be an issue with the prompt? Additionally, would using RAG to index the entire codebase help ensure the review considers the complete context? Or is there a better approach? Here is the prompt I am using: https://gist.github.com/alwin-augustin-dev/a6448f53a7577e8e1614bee361aacec0