The hard part of people analytics is rarely the dashboard—it’s reconciling disagreements between source systems so the numbers can survive scrutiny. I build the layer underneath the dashboard and make every downstream report inherit data quality instead of re-deriving it. I reconcile before I build, document metric definitions for auditable handover, and I’m honest about tooling gaps. My work is focused on metrics that drive decisions, from unifying disconnected systems to improving attribution accuracy, payment funnel insights, and reducing operational cycle times.…

Elizabeth Challenger

The hard part of people analytics is rarely the dashboard—it’s reconciling disagreements between source systems so the numbers can survive scrutiny. I build the layer underneath the dashboard and make every downstream report inherit data quality instead of re-deriving it. I reconcile before I build, document metric definitions for auditable handover, and I’m honest about tooling gaps. My work is focused on metrics that drive decisions, from unifying disconnected systems to improving attribution accuracy, payment funnel insights, and reducing operational cycle times.…

Available to hire

The hard part of people analytics is rarely the dashboard—it’s reconciling disagreements between source systems so the numbers can survive scrutiny. I build the layer underneath the dashboard and make every downstream report inherit data quality instead of re-deriving it.

I reconcile before I build, document metric definitions for auditable handover, and I’m honest about tooling gaps. My work is focused on metrics that drive decisions, from unifying disconnected systems to improving attribution accuracy, payment funnel insights, and reducing operational cycle times.

See more

Skills

Experience Level

SQL
Expert
Power BI
Expert
Data Visualization
Expert
PostgreSQL
Expert
Github
Expert
Snowflake
Expert
ETL
Intermediate
Data Science
Intermediate

Language

Work Experience

Data Engineer / Data Analyst (Extern) at Extern Inc.
July 1, 2026 - Present
Externships with Breaking Games (data engineering and analytics work) and Amazon people analytics. Focused on data reconciliation, analytics modeling, and strengthening downstream metric trust.
Data Engineer / Data Analyst (Extern) at Amazon
April 1, 2026 - June 30, 2026
Externship in Amazon people analytics. Contributed to analytics engineering and data quality improvements that support reliable reporting.
Co-Founder and Data Analyst at Creative Splash Home Improvements
January 1, 2020 - Present
Co-founded and worked as a Data Analyst supporting business decisions through analytics and reporting. Built repeatable workflows and improved how metrics were produced and interpreted.
Registered Nurse Supervisor (part-time night shift) at Justin Fadipe Medical Center
January 1, 2015 - December 31, 2017
Provided supervisory nursing leadership on the night shift while concurrently holding the role below.
Registered Nurse at Princess Margaret Hospital
January 1, 2010 - December 31, 2017
Full-time registered nurse work across 2010–2017.

Education

MBA, Healthcare Management at Western Governors University
January 11, 2030 - August 6, 2026
BS, Health and Human Services at Western Governors University
January 11, 2030 - August 6, 2026
Associate's Degree, General Nursing at Dominica State College
January 11, 2030 - August 6, 2026

Qualifications

Big SQL Energy (Jess Ramos): SQL, Hex and Snowflake | Zero To Mastery: Data and Analytics
January 11, 2030 - August 6, 2026
Excellence Award (Service Line Development) - MBA, Healthcare Management
January 11, 2030 - August 6, 2026
Excellence Award (two) - BS, Health and Human Services
January 11, 2030 - August 6, 2026

Industry Experience

Healthcare, Professional Services, Software & Internet, Financial Services, Gaming
    Subscription Cancellation Analysis
    Product Analytics · SaaS · Retention Churn had climbed past what leadership could explain, with a board review coming. I was asked why customers were leaving — using the reasons they select on their way out. First question wasn't "why do they leave" — it was "can this data be trusted?" The exit flow asks for one required reason and two optional ones. If people were clicking through to escape, the optional fields would be noise. So before any trend work, I measured engagement with binary flag logic: 81.8% selected at least one optional reason, averaging 2.18 reasons each. Reliable enough to use — but only 36.4% reached a third reason, so I treated those fields as directional, not decisive. Reasons by position and "Expensive" (31.8%) are the top primary reasons — together 68% of required selections, and the most reliable signal in the set "Went to a competitor" moves third as a primary reason (27.3%), first as a secondary one (38.9%), zero as a third. Customers cite competitors as a contributing factor, rarely as the thing that decided it "Bad customer service" is 4.5% of primary reasons — and 62.5% of third reasons. Anyone reading only the required field would file support experience as a non-issue Third reason distribution OVER (PARTITION BY year)) for year-over-year share. Recommendations delivered Onboarding investment against "Not useful" — customers who never hit the aha moment leave, and value perception blunts price sensitivity too. A rescue offer (discount, downgrade, or pause) surfaced early in the exit flow for cost-sensitive users. A competitive audit to feed roadmap priority. A support-ticket review for the eight subscriptions citing service failure. And a flow redesign — one question at a time plus optional free text — to fix the completeness drop-off the analysis exposed. Limitation: 22 cancellations. The third-reason spike rests on 8 responses, and 2022 holds a single cancellation, so annual trend is not usable. The direction is clear; the precision isn't. SQL used: CTEs · CASE · UNION · window functions (SUM OVER PARTITION BY) · view creation · DATE_TRUNC · COUNT DISTINCT · null handling Stack: SQL · Snowflake · Hex · data visualisation 🔗 View GitHub Repo: https://www.twine.net/signin…
    Subscription Payment Funnel Analysis
    Product Analytics · SaaS · FinTech The finance team had a revenue leak they couldn't locate: customers were signing up for paid plans and never paying. Because a customer counts as active at signup but converted only at payment, the company was absorbing onboarding cost for accounts generating nothing. I mapped the end-to-end payment flow with the product manager, confirmed event coverage with the frontend and data engineering teams, then built the funnel in SQL against Snowflake and analysed it in a Hex notebook. The hard part wasn't the funnel — it was the retries Subscriptions don't move cleanly forward. They hit an error, cycle back, and retry. MAX(status_id) alone credits the furthest point reached, but can't tell a subscription cleanly sitting at "payment success" from one stuck in a vendor error after it. Classification required combining max_status with current_payment_status in a CASE statement across eight stages. The other decision that mattered: a LEFT JOIN onto the full subscriptions table. An inner join would have silently dropped the 24 subscriptions with no payment-log record — deleting the single biggest finding from the analysis before it started. Results Subscriptions by funnel stage, split across two distinct failure points: user-side submission errors and vendor-side processing errors Error share · LEFT JOIN / anti-join · MAX · DISTINCT · DATE_TRUNC Stack: SQL · Snowflake · Hex · Python (pandas) · data visualisation 🔗 Link to GitHub: https://www.twine.net/signin…
    Early Associate Attrition Analysis — Fulfillment Center People Analytics
    Reducing 60-Day Attrition in a Fulfillment Center Consulting Project · People Analytics A high-volume fulfillment center was losing new Pickers, Stowers and Packers inside their first 60 days. Operations and HR knew the number; nobody could name the cause. I was asked to find where the onboarding-to-floor transition actually breaks — and what to change. *Three compounding failure points between hire and day 60.* Approach I built a voice-of-employee dataset of 145 public records — 139 Glassdoor reviews plus 6 "day in the life" transcripts — tagged across six friction themes, then scored sentiment with VADER (NLTK) in Python. Then EDA on theme frequency by tenure band, friction mapping tied to verbatim evidence, and a 5 Whys pass tracing symptom back to system. What the data said Physical Toll & Bodily Strain is the only theme carrying net-negative sentiment — 34% negative vs 10% positive In the priority segment it is the loudest signal: 39% of all mentions, 42% negative 43% of mentions in the under-1-year band are rate pressure — the dominant early-tenure driver 75% of friction mentions (86 of 114) sit at two years of tenure or less *"Maintaining that speed and quality was really draining … I ended up with a repetitive motion injury in my hand."* — Glassdoor Three root causes surfaced: a physical conditioning gap (onboarding never replicates full-rate load), tenure-blind scheduling (week one is staffed identically to year one), and metric invisibility (early attrition cost appears in no KPI a floor manager owns). Recommendations delivered A 60-day tiered UPH ramp — 60-70%, then 75-85%, then 90-100% with full-rate certification — anchored to a Mid shift whose 12pm-2pm overlap creates a supervised conditioning block at zero added headcount, modelled on nursing preceptorship. Two workforce-management fixes: a ramp coefficient so a ramping associate reads as planned capacity rather than a fill-rate miss, and a 60-day retention metric on the floor-manager dashboard. A peer-run strain check-in, and a scoped pilot (Pickers and Stowers, one pick-tower zone) before scaling. Limitation: first-60-day leavers rarely write reviews, so physical toll is structurally undercounted. The finding is directionally strong, not a census. Tools: Python · pandas · NLTK/VADER · matplotlib · Google Colab · D3.js · Google Sheets Skills: Sentiment analysis · EDA · data wrangling & cleaning · qualitative coding · friction mapping · 5 Whys root cause analysis · data visualisation · executive communication (Pyramid Principle / SCQA) All data from public sources. No internal company data was used; all identifying information anonymised. 📊 View the final deck: https://www.twine.net/signin 🔗 View GitHub Repo: https://www.twine.net/signin…

Hire a Data Analyst

We have the best data analyst experts on Twine. Hire a data analyst in Knoxville today.