Senior Data Engineer

• Senior Data Engineer — Data Warehouse & Platform
• Trunk Space Storage | Preferable Dallas, TX; Remote (US)possible

• About Us
• Baranof Holdings has been exclusively focused on self-storage since 2015, with roughly $2 billion in asset value across more than 90 properties in 21 states. Trunk Space Storage is our self-managed operating platform, running 22 of those assets directly; the rest are managed by third-party operators, principally Public Storage and Extra Space Storage.

• The Role
• We've built a Bronze/Silver/Gold data warehouse on AWS — Python, Lambda, DuckDB, and Delta tables in S3 — and it's what the business runs on: occupancy, rate and ECRI performance, delinquency, leads and conversion, and same-store revenue all come out of it. What it hasn't had is a permanent owner; contractors and borrowed capacity have carried it this far.

This role owns it. You'll make sure all the data reaches the warehouse, that the warehouse is well-structured and secure, and that it grows into a genuine enterprise capability. You're not expected to be a self-storage expert, a real estate analyst, a machine-learning engineer, or an infrastructure architect — Operations and Revenue Management define what the numbers mean, our data scientist owns statistical modeling, and an external vendor delivers our Terraform. Your job is the data: getting it in, making it consistent, keeping it correct, and putting it in front of people. External engineering vendors add capacity under your direction — you hold the expertise, they provide the throughput.

• What You'll Own
• • Ingestion across every source: our own ERP (SSM), a third-party operator API, operator file exports, and the systems around them
• • The silver layer, conforming very different source shapes into one consistent structure — and documenting clearly where they can't be
• • The gold marts: occupancy, rate and ECRI, delinquency, leads and conversion, same-store revenue
• • Metric definitions as versioned, documented code, once the business has defined them
• • Validation gates at ingestion and at gold write, reconciled against source of record
• • The Shiny reporting layer, built directly with the people who use it
• • The AWS platform underneath it all: monitoring, alerting, logging, secrets, IAM, CI/CD, and the inherited Terraform estate
• • Specifying and reviewing work sent to external vendors

Roughly: 45% warehouse (ingestion/silver/gold), 30% reporting, 20% platform reliability & security, 5% standards and vendor management.

• What You Won't Own
• Metric and business definitions (Operations/Revenue Management), operator data access (Asset Management), statistical/ML modeling (our data scientist), or authoring infrastructure as code (our Terraform vendor — you inherit, operate, and extend it).

• What We're Looking For
• • Ships production Python independently — services, scheduled jobs, error handling, logging, deployment, not notebooks
• • Has personally designed, shipped, and evolved dimensional models without breaking downstream consumers
• • Experience integrating sources you don't control: mixed APIs and file exports, different auth models, formats that change without warning
• • Judgment about correctness — you've owned a number someone made a decision on, and can explain how you knew it was right
• • Comfortable representing partial or missing data rather than filling gaps with something plausible
• • Working AWS depth: S3, Lambda, Athena/Glue, CloudWatch, CloudTrail, Secrets Manager, and writing your own IAM policies
• • Has maintained infrastructure as code you didn't author
• • Has run a platform other people depended on — your monitoring, your alerts, your incidents
• • Has shipped analytical interfaces non-technical people actually adopted (Shiny, Streamlit, Dash, Tableau, Power BI, Looker, or your own build)
• • Strong SQL: window functions, incremental/merge patterns, real discipline about grain
• • Experience with lakehouse table formats (Delta Lake or Iceberg) and an in-process engine like DuckDB or Polars
• • Comfortable turning a two-sentence business question into a scoped deliverable
• • Fluent working alongside AI coding agents, including reviewing code you didn't write
• • 5+ years building and operating production data systems

No self-storage, real estate, or machine-learning background required — we'll teach the domain, and our data scientist owns ML.

• Nice to Have
• Real estate or multi-site field-operations data experience; work with third-party/partner-supplied reporting data; Shiny (or any reactive framework — learnable here); R beyond dashboards, including the paws SDK; dbt familiarity; pytest/moto/localstack; AWS security tooling (GuardDuty, CloudTrail analysis, Lake Formation); Athena workgroup/cost governance; Terraform authoring experience; prior experience as the internal counterpart to an outsourced engineering vendor.

Not part of our stack: Kubernetes, Spark, Snowflake, Databricks, streaming/Kafka. Our stack is deliberately small and staying that way.

• Compensation & Location
• $155,000–$185,000 base, depending on experience, plus performance bonus. Remote within the United States, with periodic travel for team and site visits. Geographic adjustment for high-cost metro markets. Health, dental, and vision coverage; retirement plan; paid time off.

• Why This Role
• You own the warehouse outright, with a direct line to the executive who sponsors it — no committee, no incumbent architecture to argue with. The data problem itself is genuinely interesting: three source shapes across roughly 100 assets and two national operators, conformed into something people can trust. The work is visible — numbers that institutional capital partners read come out of what you build, and when a district manager changes how they price or collect, it's because something you shipped told them to. Direction here is set internally, and vendors execute against it: what you build becomes how we build.