Data Scientist (Remote)

Data Scientist (Remote) – Taskify AI We are looking for a data-driven professional to join our analytics and quality team. In this role, you will analyze the performance of AI agents across finance-sector tasks to uncover root causes of failure and drive improvements in design, evaluation, and benchmarking frameworks.

Key Responsibilities

Conduct statistical failure analysis to identify recurring patterns in AI agent performance across prompts, rubrics, templates, and file types

Investigate root causes behind performance issues in task execution, framework design, or data complexity

Perform multi-dimensional analysis across finance sub-domains, task categories, and input types

Develop dashboards and reports highlighting performance trends, failure clusters, and improvement opportunities

Recommend enhancements to task design, quality rubrics, and evaluation metrics based on data findings

Required Qualifications

Strong understanding of statistical analysis, hypothesis testing, and pattern recognition

Proficiency in Python (pandas, scipy, matplotlib / seaborn) or R for data analytics

Hands‑on experience in exploratory data analysis (EDA) and turning data into actionable insights

Familiarity with AI / ML evaluation techniques and quality assessment metrics

Proficiency in Excel, SQL, and visualization tools such as Tableau or Looker

Preferred Qualifications

Prior experience in AI / ML model evaluation, data labeling quality, or automation analysis

Knowledge of finance-sector concepts or willingness to learn finance domain structures

Experience with multidimensional failure analysis or benchmarking studies

Understanding of large-scale evaluation frameworks or test dataset design

Equal Opportunity Statement :

We welcome applicants from all backgrounds. Reasonable accommodations are available for qualified individuals with disabilities.

Seniority level : Not Applicable

Employment type : Part-time

Job function :

Information Technology

Industries :

IT System Data Services and Technology, Information and Media

#J-18808-Ljbffr