Remote (Eastern Time Zone)

Sophie Searcy

| s@soph.info | soph.info


Summary

Sophie is a Machine Learning leader who grows engaged, productive teams and builds impactful Data Science and Machine Learning products. Her no-nonsense approach to ML and leadership prioritizes trust, clarity, and action. Python (expert), Javascript (fluent), R (fluent), SQL (fluent) .

Experience

Stitch Fix

Acting Director of Merch Core Models
2021-
Remote

Leads group that includes 2 teams and 8 ICs charged with building and maintaining the foundational models that drive merchandising decisions and forecasts for our $2 Billion retail business and 3.4 million customers.

New: Launched two products that have brought our Algos expertise to bear on decisions for new products for the first time. Design The Line is a creative tool that supercharges the Buyer-Vendor relationship by providing predictions of business metrics for products throughout the design of and decision-making on new products. Merch Attribute Extraction uses State of the Art computer vision models to remove the distinction between new and known products and dramatically extend the leverage of our buying decisions.
Forecasting: Led development of performance monitoring for Allocation and Replenishment. Reduced maintenance burden of internal sales forecast capability by 50% while protecting against adverse business impact.
Leadership: Has grown and retained a team of 8 high-performing ICs, delivering on challenging, business-critical work, with zero attrition during a period where internal DS attrition has approached 50% YoY.

Metis

Executive Director of DS, Principal Data Scientist
2017-2021
New York, NY

Data Science education company whose main product is a 12 week bootcamp that covers the breadth of the field of Data Science. Current role (>2 years) is primarily EM & PM.

Product: Oversaw the design and development new and growth products across org. Led cross-functional product teams of >15 individuals. Current product portfolio represents 25% of 2021 revenue and accounts for 100% of growth opportunities.
Management: Directly manages team 6 Data Scientists. Introduced industry best practices including company-wide Eng and DS prioritization flow, code reviews, and continuous integration.
Instruction: Taught and mentored 100s of students entering data science. Authored dozens of lessons including: SVMs, Data Science Ethics, Deep Learning, Generative Adversarial Networks.

Elektra

CTO, Cofounder
2016-2017
New York, NY

Wearables start-up building fabrics that pair electrical stimulation to apps. Fundraising: \$1.5M cap. Competitions: The Bridge (Coca Cola’s Accelerator) finalist (ranked 22/200). AlphaLab Gear National Hardware Cup semi-finalist. Vogt Awards finalist (\$25k non-dilutive funding). XlerateHealth finalist. Bucks for Bright Ideas winner.

Proof of concept: Designed and built hardware and software proof of concept. Validated underlying technology, collected data on 100+ customer experiences in target audience and demonstrated product-market fit.
Scalable design: Led engineering teams managing user data and designing a product for 10k/yr volume: the entirely in-house product stack included electronics hardware, firmware, and desktop app.
Human resources: Managed HR. Designed collaborative, version controlled, company Handbook.

Cognitive and Data Science Lab

Research Assistant, PhD Candidate
2012-2017
Newark, NJ

Project manager: Managed a data science team developing FOSS tools for replicable cross-platform video data processing at “Big Data” scale.
Web experiments: Designed full-stack web apps for large (n~1,000) behavioral experiments and administered experiments through Amazon Mechanical Turk. Mentored data scientists in full-stack design.
Interdisciplinary research: Lead author of interdisciplinary work on the foundations of learning and computational tractability of learning algorithms. Presented work in journals and conferences.

GE Appliances

Electronics Engineer
2012-2017
Louisville, KY

Talks and Publications

Searcy, Burgess (2022) Empowering Shopping with Computer Vision

Searcy (2019) AI is Broken

Searcy (2019) Building Generative Adversarial Networks in Tensorflow and Keras

Searcy and Blinder (2018) Making Faces: Conditional generation of faces using GANs via Keras+Tensorflow

Searcy, W.K. Vong, Lane & Shafto (In prep) Cause and affect: Subjectively rational causal inference about affective content.

Searcy (In prep) Levels of expressivity: Learning across incompatible representational levels

Searcy & Shafto (2016) Cooperative Inference: Features, objects, and collections. Psychological Review.

Searcy & Shafto (2014) Learning biases for teaching boolean concepts.

(2010) Energy Efficiency Comparisons of Wireless Communication Technology Options for Smart Grid Enabled Devices

Education

PhD (ABD), MS

University of Louisville | Experimental Psychology: Computational Cognitive Science

MEng, BS Eng

University of Louisville | Electrical and Computer Engineering
University Fellowship Award (2 years Full tuition plus stipend), National Merit Scholarship (5 years full tuition plus stipend)