Indonesian Job-Listings NLP Labeling & Enrichment (internal platform project)
Built and maintained labeled Indonesian-language datasets powering lokerdollar.com's job-matching platform. Work included: classifying and normalizing job postings (role, seniority, remote/on-site, salary band); named-entity extraction of skills, companies, and locations from Bahasa Indonesia text; and curating prompt/response data for a Bahasa job-search chatbot. Applied rubric-driven labeling with gold-standard checks and review passes to keep label quality consistent across a large, continuously ingested listing feed. Demonstrates native Bahasa text classification, NER, and conversational-data curation at scale on our own production system.