Data Scientist
Streamlit, Plotly, numpy, pandas, matplotlib, seaborn and with AWS - S3 a plus. Someone who is smart and truly a data scientist.
Platform: Python/Streamlit production intelligence platform for the Amex Travel Mobile App — centralizes telemetry (Snowflake, ELF/OpenSearch, App Store Connect, Jira, etc.) into dashboards for TLS leadership. Strict 4-layer architecture: Ingestion → Analysis → S3 Storage → Visualization.
Technical Skills
Strong Python 3.11+, pandas, NumPy for data transformation and time-series work
Streamlit dashboard development (multi-page, session state, query-param nav)
API/SQL data integration (Snowflake, OpenSearch, REST sources)
Visualization: plotly, matplotlib, seaborn
AWS (boto3, S3/Parquet), Docker, Kubernetes, CI/CD
Secrets/auth (JWT, cryptography, proxy handling); testing with pytest
Data Science / Analytics
Applied statistics for anomaly detection (rolling median + MAD z-score, thresholds); scipy a plus
Time-series thinking: daily/weekly/monthly frequency-aware baselining
Turns raw telemetry into decision-ready insights (not deep ML, but solid quantitative reasoning)
Business Knowledge
Understands mobile app KPIs across Bookings, Adoption, Engagement, App Health, Card Member Voice, Quality Engineering
Familiar with iOS/Android store metrics, crash/stability, QE/regression concepts
Translates leadership questions into metrics and views
Seniority
Senior-level ownership within an established layered architecture; respects design principles (no business logic in UI, filter-agnostic analysis)
Contributes to and follows Architecture Decision Records (ADRs)
Owns features end-to-end: data source → analysis → dashboard
Communication & Collaboration
Communicates data-driven findings clearly to leadership and non-technical stakeholders
Works cross-functionally with QE, delivery, and product teams
Strong documentation habits
Go-Getter
Self-directed in integrating new data sources and extending hubs/domains
Curious and metric-obsessed; proactively closes visibility gaps
Bias to ship while keeping the architecture clean
