Ask HN: Who is hiring? (March 2025)
Role types & domains
- Wide variety of senior engineering roles: backend, full-stack, infrastructure, SRE/DevOps, data/ML, security, and product engineers.
- Many AI- and LLM-related positions: agent frameworks, model training/inference infra, applied AI in domains like healthcare, legal, marketing, and developer tools.
- Strong presence of fintech, climate/energy, industrial/robotics, devtools, security, and healthtech startups, plus some larger established companies.
- Non-engineering roles include product managers, data scientists, growth marketers, developer advocates, UX/design, sales, and operations.
Remote, location & visas
- Many jobs are nominally “remote” but restricted by region (e.g., US-only, EU-only, specific time zones) or by state/country employment constraints.
- Frequent questions about visa sponsorship and relocation; answers vary widely: some explicitly sponsor (esp. EU), others cannot hire outside specific jurisdictions.
- Several on-site or hybrid-only roles in NYC, SF Bay Area, London, Berlin, Amsterdam, etc., with 2–5 days/week in-office expectations.
Hiring practices & candidate experience
- Some posts highlight lean, practical interview processes (few rounds, work-sample or real-world tasks; explicit rejection response promises).
- Others draw criticism: long take-home assignments without feedback, generic rejections after multiple stages, or no response at all.
- One commenter notes concern about “perpetually open” postings and mentions an informal community push to discourage evergreen reqs.
Trust, reputation, and meta-discussion
- A few companies are accused of “fake openings” or rejecting experienced candidates without interviews; one responds citing active hiring numbers.
- Detailed complaint about a firm allegedly extracting unpaid consulting ideas before canceling a project; other commenters encourage re-posting as a warning.
- A candidate shares frustration with lack of feedback from a company that promotes transparency.
- Separate thread explores prompt-injection-style instructions inside a posting and how current LLMs often still obey them, despite specs saying they shouldn’t.