Skip to main content
Shameek ConyersData ScienceResumeContact

For Hiring Teams

Data scientist since 2021, open to a full-time role.

Shameek Conyers

In short

I'm a data scientist looking for a full-time role. My strongest work is with marketing and sales data: measuring what each channel brings in, forecasting demand, and deciding which leads to call first. I'm also open to applied data science roles outside marketing, such as machine learning models, forecasts for staffing and stock, and AI tools built on a company's own documents.

  • Experience: 4+ years as a data scientist, since November 2021. Data Scientist at Ringer Sciences, a marketing analytics consultancy, from 2021 to 2024, on a small analytics team reporting to the head data scientist. Independent consulting since 2024.
  • Working alone: I've been the only data person at a business. I set up the tracking, built the reports, and reported to the owner directly.
  • Education: B.A. in Mathematics and B.S. in Econometrics and Quantitative Economics, University of South Florida.
  • Location: U.S. citizen in Tampa, on East Coast hours. Open to remote work or to relocating, with New York preferred.

View my resume or email me about a role.

What I'd do on your team

I'd start the way I start with clients:

  1. Agree with sales and marketing on what counts as a lead, a sale, and a customer, and which decisions the numbers need to support.
  2. Connect the records you already keep, such as the ad accounts, the CRM, and the sales system, and find where they don't line up.
  3. Report which spend brings in revenue, in one view the team can check every week.
  4. Then add what the team needs next: forecasts of what a budget change should do, and a fair comparison to check whether it did.

I decide what to compare against before I look at the numbers, so the result can't be bent to fit a story.

How I work with people

At one cafe client I was the only analytics and marketing person. I reported to the owner and briefed the creative contractors and directed their work. When the owners disagreed, I settled it with a test and a range of likely outcomes instead of an opinion. When a new owner wanted to spend more on ads, I used visit forecasts to show the staff and stock the extra visits would need, and that tempered the plan.

For a wholesale supplier, the ask was vague: make the sales calls more efficient. Before building anything, I asked clarifying questions and costed several options against the current process.

At Ringer Sciences I worked across six client engagements and presented findings to client marketing leadership. For a hospital client, I built dashboards for two audiences at once: the marketers running the campaigns, and the physicians who needed the takeaways rather than the mechanics.

Before data science, I spent nearly four years in computer sales at Best Buy while studying, explaining technical products to customers.

To see how I report results to people who don't work with data, read the sample readout. It's built on a made-up business.

Selected work

  • Marketing measurement for a hospital. Dashboards for a multi-million-dollar physician marketing program, joining hand-typed journal ad records with ad platform data. The hospital used them to decide where to move its marketing budget and kept paying for them.
  • Campaigns tied to sales. For a cafe and lounge, the ad spend behind each dollar of tracked sales fell about 70% on lead campaigns, after I rebuilt the ad audiences from customers who actually walked in.
  • Lead ranking. For a wholesale supplier, closing rate per call rose about 50%, from about 8 closes per 100 calls to near 12, once reps called the leads my model ranked first.
  • Search over past research. At Ringer, a search assistant over a B2B media client's own reports. Over a one-quarter pilot, its marketing and PR department reported about 40% less research time, and the client kept it as a paid product.
  • Channel budgets. At Ringer, client teams used my marketing mix and forecasting models to set channel budgets.
  • Scale. Text analysis of more than 2 million social posts in daily pipelines at Ringer. Client teams changed media and content plans on the findings, and the firm sold the work to other clients.
  • Checking my own work. In a public study, I published the plan for a follow-up test before running it on 36,774 reviews the study had never used. 21 of 24 findings held up.

Each case study says how its result was measured, including when it was a before-and-after comparison rather than a controlled test.

Independent work since 2024

Since February 2024 I've worked with food and beverage, agriculture, and consumer businesses. In a small business I'm often the only analytics person, so the work runs from campaign setup to the owner's decision. One client, a cafe that became an event lounge, kept me on retainer as its marketing data scientist.

Tools

  • What your team would see: Google Ads, Meta Ads, Google Analytics, Excel, Looker, Tableau, and Power BI.
  • What I build with: SQL, BigQuery, dbt, Python, R, scikit-learn, LightGBM, PyTorch, spaCy, and Hugging Face.

Next step

Send the posting or a short note about the team to shameekconyers@gmail.com. I reply within a day.