[ Subject File / Live ]

Dumebi Gloria Akukwe

+1 845 214 2717 · Dumebiakukwe2@gmail.com · linkedin.com/in/Dumebi-Akukwe

01

Education

Marist University — Poughkeepsie, NY

Master of Science, Information Systems & Business Analytics · Dec 2025

Marist University

Bachelor of Arts, Economics · Minors: Business Administration, French · May 2022

02

Technical Skills

Languages & Tools

Python · PostgreSQL · Excel/VBA · Git · GCP (BigQuery, Cloud Storage) · Claude API · Gemini API · Tableau · Power BI

AI & Automation

Prompt Engineering · LLM Orchestration · Agentic Workflows · NLP (BERT, TF-IDF) · Machine Learning · Predictive Modeling

Data Systems

Data Pipelines · Data Quality · Multi-tenant Architecture · LLM Integration · Statistical Analysis · A/B Testing

Product

Full Product Lifecycle Ownership · Roadmap Prioritization · User Research · Stakeholder Management · SaaS Development

03

Professional Experience

Founder & AI Product Lead

HireRail.io

AI-Powered HR & Recruiting SaaS

Jun 2024 – Present

  • ·Architected and shipped a production AI screening engine end-to-end — from discovery and roadmap through engineering coordination, QA, and launch — reducing per-candidate review time by 75%.
  • ·Designed a five-dimension candidate assessment framework with temperature calibration and fairness guardrails, translating recruiter and compliance requirements into AI-powered workflows.
  • ·Built an agentic hiring workflow with multi-stage pipeline management, credit-based usage metering, and webhook-driven Stripe billing across a multi-tenant SaaS platform.
  • ·Led prompt engineering for a structured candidate evaluation system with no-fabrication constraints and bias guardrails for cross-regional talent assessment.

Product & Curriculum Architect

MsSuzie.io

K–12 EdTech Startup

Sept 2024 – Dec 2024

  • ·Developed NLP classification models analyzing student input across 30+ behavioral and learning indicators, enabling automated learner diagnostics.
  • ·Designed data collection frameworks to standardize multi-source K–12 data, improving analytical consistency.
  • ·Built rule-based decision models converting unstructured tutor–student interactions into measurable performance metrics.

Graduate Assistant — Data Strategy & Public Engagement

Marist University, Office of Admissions

Executive dashboards · Recruitment analytics

Jan 2024 – Dec 2025

  • ·Built and maintained executive dashboards in Salesforce and Excel analyzing recruitment funnel performance, reducing reporting turnaround by 40%.
  • ·Analyzed multi-source global survey data (2,800+ respondents), informing institutional strategy including Marist's transition from college to university.
  • ·Performed ad-hoc and recurring analyses supporting resource allocation and campaign decisions.

Data Analyst (Intern)

Success Academy Charter Schools

SQL pipelines · Performance analytics

Jun 2023 – Jan 2024

  • ·Engineered SQL data pipelines integrating multiple datasets, reducing reporting latency by 50% and improving accuracy.
  • ·Conducted statistical analyses on academic and behavioral datasets to evaluate interventions and identify trends.
  • ·Delivered insights and trained 60+ stakeholders, improving organization-wide data fluency.

04

Featured Projects

Fraud Detection & Investigation System

  • ·Detected fraud across 590K transactions using XGBoost (0.924 AUC) with end-to-end feature engineering: velocity, address completeness, email domain, card type.
  • ·Quantified causal fraud drivers using DoWhy and Difference-in-Differences — credit cards 4.36 pp riskier; missing billing addresses carry 5× higher fraud rate.
  • ·Built SHAP explainability layer and LLM-generated plain-English summaries so model predictions are interpretable for non-technical reviewers.

NLP Clustering & Anomaly Detection — Mental Health Signals

  • ·Analyzed 33,554 Reddit posts using TF-IDF, BERT embeddings, and K-Means clustering to identify behavioral patterns in unstructured text.
  • ·Detected hidden high-risk signals — ~26.5% of posts contained crisis indicators missed by baseline sentiment models.
  • ·Applied association rule mining (428K+ rules) and sentiment analysis to uncover linguistic patterns and improve interpretability.

05

Awards

PPIA Junior Summer Institute Fellow — Princeton University

Advanced Economics Track · Fully funded fellowship · 2022

Marist Presidential Scholarship

Top merit-based scholarship for excellence and impact · 2019–2022

06

Leadership & Community

  • Volleyball League Player — Mid-Hudson Athletic Center, DC Sports League · Present
  • Founder & President, African Student Association — Marist University · 2021–2022
  • Operations Director (Volunteer, Remote), Eastern Nigeria Film & Arts Initiative · 2024
  • Treasurer, Marist Graduate Student Network · 2025
  • Volunteer, Restoration Food Pantry · 350+ hours · 2019–Present