Head of Data Infrastructure

About Beacon

Beacon Software acquires and operates vertical market SaaS companies across North America. We run a portfolio of 25+ software businesses spanning campground management, election platforms, field safety tools, proptech data, workforce scheduling, and more, with a shared focus on compounding customer value over the long term.

Data sits at the centre of how we operate. We use it to run portfolio companies better, underwrite new acquisitions, and increasingly to build AI-powered capabilities into our products. This role is the person who architects, owns, and builds the infrastructure that makes all of that possible.

The Opportunity

We are looking for a senior data infrastructure leader who is still deeply hands-on. Someone who has spent time inside companies like Snowflake, Databricks, or Confluent building the platforms others run on, and who now wants to apply that depth to a genuinely interesting operational problem.

You will be the first dedicated data infrastructure hire at the Beacon holding-company level. You will design the architecture, write significant amounts of the code, and set the patterns for how data flows across a growing portfolio. The problem is not generic. Multi-tenant isolation across very different portcos, cross-cloud ingestion from systems we do not control, regulated verticals with residency requirements, and a long-term path toward making this data usable by both people and software. Over time, you will define whether and how a small team gets built around you.

What You'll Own

• The central data platform. Design and build Beacon's lakehouse, warehouse, and pipeline architecture. Unify data across 25+ portfolio companies into a coherent platform that survives contact with very different source systems, very different data quality, and very different portco maturity levels.

• Ingestion strategy. Connect diverse source systems (SaaS apps, transactional DBs, Stripe, Salesforce, custom APIs) into a low-latency data layer. The hard part is not any individual connector. It is the operating model that makes onboarding portco 50 as fast as portco 5.

• Modeling and transformation standards. Define the data modeling standards, transformation logic (dbt, Spark, or equivalent), and lineage practices used by the entire portfolio. The canonical model is the thing that lets a query like “show me sales across all portcos” actually resolve, and it is the thing that determines whether the platform compounds or fragments as we grow.

• Real-time and batch frameworks. Establish the processing framework for operational analytics, AI feature engineering, and executive reporting. Batch where batch is right. Streaming where it matters. The discipline to know which is which.

• Multi-tenant isolation and security posture. Per-portco data, compute, and credential boundaries. Cross-cloud (AWS and Azure) connectivity. Regional residency for portcos in regulated verticals. KMS scoping, IAM design, and the audit surfaces that make it all defensible.

• Infrastructure-as-code for the data stack. Cloud provisioning, cost management, and observability. You will own the Terraform or Pulumi modules that other engineers build against. The bar is not novelty. It is correctness, repeatability, and the kind of module design that ages well.

• AI and feature serving partnership. Partner with AI and product teams to build feature stores, vector infrastructure, and low-latency serving layers as we embed intelligence in portfolio products. You do not need to be the ML expert. You need to be the person who makes sure the data layer underneath is ready when ML needs it.

• Vendor stack ownership. Evaluate and own the cloud warehouse, orchestration, cataloging, and BI tooling. Build vs. buy decisions, with long-term leverage in mind. We have an active evaluation underway between Snowflake and Databricks, and you will be the senior voice in landing that.

• Acquisition diligence. Act as the senior technical voice in acquisition diligence, assessing target companies' data maturity and integration complexity. Some of the highest-leverage moments in our operating model happen pre-close, and the data infra perspective is missing from most diligence processes.

Who You Are

Technical depth

• Hands-on background in distributed systems, query engines, or cloud data platforms.

• Comfortable writing production Python, SQL, and infrastructure-as-code (Terraform, Pulumi).

• Deep familiarity with modern lakehouse and warehouse internals (Iceberg, Delta, Parquet, Snowflake,