Scott Koskoski
Sr. Analytics Engineer | Data Scientist | ML Engineer | Python, AI, SQL | M.S. Computer Science ’26
- Role
- Senior Analytics Engineer at Green Mountain
- Location
- Memphis, TN, US
- LinkedIn followers
- 500 followers
About Scott Koskoski
I build machine learning systems that run in production — not just notebooks. Over 6+ years across logistics, retail, and fintech, I\'ve taken models from research to deployment: GPU-accelerated freight optimization on 1M+ shipments, a real-time fraud detection platform processing 21K+ TPS through Kafka with XGBoost scoring at sub-10ms latency, and an internal RAG chatbot grounding a 120B-parameter LLM in company knowledge via TF-IDF retrieval.Currently a Senior Analytics Engineer at Green Mountain Technology, I work daily in Python, SQL, and cloud infrastructure to ship end-to-end ML products. My sweet spot is the full pipeline — data engineering, model training (PyTorch, XGBoost, scikit-learn), MLOps (drift detection, automated retraining, model versioning), and serving via FastAPI or streaming architectures. I write the React/TypeScript frontend when the project needs one, but my focus is on the ML systems underneath.What sets me apart: I pair deep ML engineering with an economics background (M.A, University of Memphis) that keeps me anchored to business impact. Every system I build ties back to a measurable outcome — dollars saved, hours eliminated, decisions accelerated.Currently finishing my M.S. in Computer Science at UT Knoxville (Data Mining & Intelligent Systems, 2026). Always interested in connecting with teams building production AI/ML infrastructure.Explore my code & projects: https://github.com/scottkoskoski
Experience
Senior Analytics Engineer
Nov 2025 — Present · Memphis, TN, US
Scaled a SQL-based pricing analytics platform by designing data transformations, building rate table pipelines, and extending calculation logic to support new business entities and vendor integrations- Built an internal RAG chatbot grounding a self-hosted 120B-parameter LLM (Ollama) in a 65-page Confluence knowledge base using TF-IDF section-level retrieval with conversation-aware context boosting – zero per-query API calls, deployed via Docker Compose and Open WebUI.
Education
University of Tennessee, Knoxville
Master of Science - MS, Computer Science
2024 — 2026
University of Memphis
Master of Arts - MA, Economics
2017 — 2019
University of Arkansas
Bachelor of Arts, Economics
2015 — 2017
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.