Alan Mathew

Indonesian native speaker based in Indonesia, using Indonesian (Indonesia) daily in professional and personal communication, with strong reading and writing comprehension in English. Detail-oriented and comfortable working with structured guidelines in remote, project-based environments that require accuracy and consistency. Experienced in ads rating, maps rating, image relevance annotation, and rationale-based review of AI-generated responses.

Has a background in high-volume operational data entry, validation, and structured reporting from on-site administrative work. Skilled in applying detailed guidelines, maintaining accuracy and consistency, and delivering high-quality results across data evaluation and documentation workflows.

AI Evaluator at Data Annotation (2025-12 – Present)

• Prompt Engineering & Design: Designed and executed complex, multi-turn conversational prompts to evaluate the model's ability to integrate personal context.
• Analytical Assessment: Evaluated AI performance across key dimensions: Grounding, Integration, Personalization, Content Completeness, Collaborativity, Writing Style, etc.
• Quality Assurance & Evaluation: Conducted rigorous Side-by-Side model comparisons, and ranking responses based on intent fulfillment, helpfulness, personalization, user experience, and across key dimensions.
• Data Annotation: Identified and documented technical flaws such as hallucinations, 'overnarrating,' and forced connections in AI-generated responses.
• Technical Documentation: Authored clear, structured, and defensible rationales for model rankings, providing detailed feedback to improve the model's reasoning and multilinguality (Indonesian-English).
• System Verification: Extracted and verified 'Debug Info' to ensure chat summaries and diverse data sources were utilized correctly by the AI architecture.
• Performed image relevance annotation to evaluate alignment between images and prompts.

Evaluator at Peroptyx (2026-01 – Present)

• Rated map search results to improve location accuracy and user navigation experience.
• Validated geospatial data using structured guidelines and real-world context.
• Analyzed and graded verbal interactions based on specific linguistic criteria, including intent accuracy, fluency, and context relevance.
• Provided structured feedback and quantitative scores for voice datasets to support the development of high-quality audio models.
• Identified and documented acoustic anomalies, ensuring all outputs met rigorous project specifications.

Ads Quality Rater at WeLocalize (2026-01 – Present)

• Performed ad relevance and quality evaluations to improve ad targeting accuracy.
• Applied structured guidelines to assess ad intent, trustworthiness, and user value.

Administration Staff at PT Daliatex Kusuma (2022-02 – 2025-03)

Responsible for administrative and reporting functions to support operational activities in the knitting department, which is responsible for producing greige fabric, including data processing, report preparation, and cross-department coordination.

• Performed daily high-volume data entry related to production, materials usage, and operational records with strong emphasis on accuracy and consistency.
• Validated and cross-checked operational data before reporting to ensure completeness and data quality.
• Maintained structured datasets and documentation used for daily, weekly, and monthly operational reports.
• Applied spreadsheet formulas and basic automation to reduce manual errors and improve data reliability.
• Organized digital records, reports, and backups to ensure data integrity and traceability.

Bachelor's Degree in Information Systems – Bina Nusantara University (2022 – 2024)