📢 We are looking for an engineer who designs and develops RAG pipelines based on large-scale documents including financial disclosure data, and is responsible for performance evaluation and operational stability.
1. Vision
We believe that AI will free humanity.
Lucy innovatively improves the data which is the core of that AI. Great datasets are required for AI to solve problems better in various fields such as finance, medicine, and law.
Data that enhances LLM performance is essential, and data that reduces computational power to meet the growing AI demand is necessary, aiming to ultimately advance human life.
2. What we do - Proprietary Technology
To become the monopoly mentioned by Peter Thiel, proprietary technology is the most important factor.
We develop proprietary technology that transforms data structures yielding optimal performance in AI from traditional JSON structures to other forms.
3. If you believe you can achieve great results, please email 'help@ezar.co.kr' explaining why you applied to our company and what you believe you can do well. Only those who send emails will proceed with the hiring process.
• Design and develop RAG pipeline
- Implement the entire process from document collection → preprocessing → chunking → indexing → search/re-ranking → generation
• Enhance data structure and chunking strategy
- Improve metadata/schema/chunking strategies tailored to the characteristics of unstructured documents
• Design vector DB and tune search performance
- Optimize search quality and speed through indexing/parameter/filtering/hybrid search
• Design RAG evaluation metrics and build automated evaluation pipeline
- Design quantitative metrics (accuracy, latency, etc.) and automate testing
• Develop RAG-based LLM services
- Develop content such as summary reports and visualizations
• Improve LLM response quality and enhance prompts
- Enhance prompts for improving response quality, such as reducing hallucination and citing sources
• Other related tasks
• More than 2 years of experience in AI development with Python
• Master's degree or higher
• Experience in designing and operating RAG systems
• Experience in developing ML/DL models
• Experience in developing and operating RAG-based LLM services
• Experience with financial disclosure data (SEC, DART, etc.) or experience in processing financial domain documents
• Experience in processing large datasets, designing REST APIs, and developing servers
• Experience publishing in top-tier AI/Data Science conferences/journals
• Strong understanding and problem-solving abilities in AI, RAG, and Data Science
👍 This experience particularly matches well
• Experience in measuring and improving search quality based on quantitative metrics (e.g., recall@k, mAP, nDCG, etc.)
• Experience solving RAG errors caused by document structure (tables/footnotes/images/attachments) through data design
• Experience designing and operating a QA system that monitors and improves issues in operational environments
Benefits and Work Environment
1. Service
https://lucydata.ai
2. Goals
Grow into an AI data supply company for global financial institutions in the US and Singapore by 2026
3. Culture
• A place where professionals work autonomously.
• No internal politics. Only evaluated by the output
• Leaders lead by example
• Fair evaluations and corresponding rewards
• Easy to talk. When evaluating people, do so by 'action' and 'result', not words
• If unable to show results, will have to leave the team.
1st video interview
2nd technical interview
