Edgar Erlangga

Full-stack developer and AI automation practitioner with hands-on experience building and maintaining web applications, business platforms, APIs, databases, and production workflows. Skilled in Python-based data processing, computer vision and video pipelines, AI-assisted automation, debugging, validation, deployment, and quality control. Highly experienced in LLM model evaluation (RLHF), strict instruction constraint auditing, and detailed AI data annotation using CVAT and Supervisely, with strong attention to linguistic consistency, edge cases, and remote collaboration.

AI Data Trainer / Indonesian Language Expert (Project Aether) at OUTLIERAI (2026 – Present)

• Evaluated and annotated Large Language Model (LLM) outputs to ensure natural language fluency, grammatical accuracy, and strict compliance with Indonesian EYD V standards.
• Performed rigorous constraint auditing (zero-tolerance policy) to verify model adherence to complex instructions, including negative constraints, structural limits, and formal/informal register nuances.
• Executed clean verbatim formatting and precise code-switching (Indonesian-English) annotations to refine model transcription and context-switching capabilities.
• Analyzed semantic structures and idiomatic expressions to correct literal translation biases, logic failures, and AI hallucinations.

Full-Stack Developer at JAGAT SEJAHTERA ABADI LLC (Present)

• Built and maintained full-stack platforms for real estate, e-commerce, education, services, and digital-marketing use cases.
• Developed a real-estate portal connecting agents, property listings, and prospective buyers, covering frontend, backend, database, API integration, debugging, and deployment.
• Created an AI customer-service workflow to support responses and management of prospective real-estate clients.
• Built the Kodeko e-commerce and ordering platform with product catalog, cart and order flow, and administrative management.
• Designed the Montes Academy learning portal and contributed to the One Pro Jasa service-business platform.
• Managed and improved production business websites, diagnosing issues and implementing maintainable fixes across the application stack.

Independent Project at AI DATA ANNOTATION & COMPUTER-VISION PRACTICE (2026 – Present)

• Completed video and image annotation workflows in CVAT and Supervisely using bounding boxes, polygons, masks, segmentation, keypoints, object tracking, keyframes, and interpolation.
• Performed frame-level and temporal quality control, including tracking-drift correction, ID-continuity checks, missing-object review, occlusion and off-screen handling, and edge-case documentation.
• Compared manual, interpolation, and AI-assisted workflows to improve annotation efficiency while preserving label accuracy and consistency.

Vocational High School in Sistem Informatika, Jaringan, dan Aplikasi – SMKN9 Kota Bekasi (2021 – 2025)