The Machine Learning team at JPMorgan Chase applies cutting-edge machine learning techniques to the firm’s unique data assets to improve decision-making across lines of business.
As a Data Domain Architect Analyst in the Annotation Center of Excellence (ACoE), you will deliver high-quality annotations across text, chat, email, document, and audio data, producing validated ground-truth datasets for ML/GenAI use cases. You will manage end-to-end annotation work—preparing data, labeling entities and relationships, maintaining taxonomies and guidelines, applying prompt engineering for consistent workflows, and tracking quality metrics—while partnering with ML/data teams and using Python as needed for data prep and QA.
Job Responsibilities :
- Own end-to-end annotation engagements: understand business objectives, identify relevant data, annotate, validate, and deliver high-quality ground truth datasets.
- Use data labeling tools to annotate structured and unstructured data (e.g., chats, emails, documents, and audio) with high accuracy and consistency. Conduct entity and relationship labeling (including entity linking/disambiguation) and define relationships among entities based on business context.
- Apply prompt engineering to design/test/refine prompts supporting consistent labeling and evaluation workflows.
- Apply financial domain knowledge to interpret language nuances and ensure annotations align to agreed business definitions.
- Should be conversant with semantic labeling, define and apply consistent label taxonomies, sub-labels, and hierarchical labeling schemes aligned to use cases.
- Transcribe and annotate audio (single/multi-speaker) across dialects/accents, including keyword, intent, and sentiment labeling where required.
- Establish and maintain annotation guidelines and standards; evolve them as requirements change.
- Validate model outputs from a business perspective and provide structured feedback for model improvement. Define and track annotation quality metrics and operational KPIs; drive continuous improvement.
- Use Python scripting for data preparation, sampling, preprocessing, and QA/audit checks (as applicable).
Required qualifications, capabilities, and skills
- 2+ years of post qualification professional experience.
- MBA or Master’s in finance and/or Data Analytics discipline
- Excellent written and oral communication skills to clearly present analytical findings and business recommendations. Highly motivated, productive, and teamwork oriented.
- Working-level understanding of ML concepts and evaluation metrics (precision, recall, F1-score) and how data quality impacts model outcomes.
- Strong attention to detail; ability to work independently and collaboratively in cross-functional teams.
- Strong financial domain knowledge and ability to interpret financial language in text and speech.
- Experience extracting/collecting data from financial documents and unstructured sources (emails, reports, chat logs).
Preferred qualifications, capabilities, and skills
- Familiarity with industry-standard annotation approaches and quality frameworks. Experience with annotation concepts/methodologies (guidelines, QA, taxonomy management, agreement/consistency practices).