SYSTEM ARCHITECTURE
Agentic workflow that sources internship and job listings, filters by visa requirements, and prepares semi-automated applications.
A student-focused career services product needed to surface internship listings matching very specific criteria (visa sponsorship, internship-only, posted within 7 days, specific role types).
The existing approach was manual filtering through job boards. Students spent 5+ hours per week sorting through irrelevant postings before they could even apply.
The team needed an agentic system that sources from multiple boards and company career pages, then aggressively filters using LLM reasoning about visa sponsorship language and role suitability.
Playwright crawlers for LinkedIn, Indeed, Glassdoor, Workday, Greenhouse, Lever.
GPT-4o evaluates each posting against strict criteria.
Structured logic with LLM fallback for ambiguous postings.
Populated daily with passing roles, ranked by match quality.
New matches with one-line summaries.
Tailored cover letter per role.
Students review every application before send.
n8n
PythonPlaywrightMake.comLangChainCrewAIGPT-4oNotionAirtableGoogle Sheets
Free 30-min scoping call. No pitch deck, no obligation, just a conversation about what's worth building.
Book a 30-min scoping call