The OpportunityWe 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 AreTechnical 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, Databricks).
- Strong grasp of streaming architectures (Kafka, Kinesis, Flink) and CDC patterns.
- Experience building for both analytical and operational workloads at scale.
Operating style
- Still gets excited to write code and review PRs. This is not a purely strategic role.
- Operates well in small-team, high-autonomy environments without heavyweight process.
- Strong opinions, loosely held. Makes architectural decisions fast and iterates.
- Can translate infrastructure complexity into plain language for operators and GMs.
- Comfortable scoping work across many different business contexts simultaneously.
What This Role Is Not- This is not a heads-down IC role with no scope. You will be setting the technical direction for data infrastructure across the entire portfolio.
- This is not a purely strategic role. We expect you to write code, review PRs, and be in the systems day-to-day. The team you build around you will follow the bar you set.
- This is not a maintenance role. The platform is net-new. You will be designing it from scratch and living with the consequences.
Compensation & LogisticsCompetitive base salary plus meaningful equity in Beacon. Comp bands are calibrated to top-of-market tech companies, not to PE firm norms. Beacon HQ is in San Francisco (CA). Hybrid expected, with the expectation that you will travel to San Francisco (CA) or Toronto (CN) regularly for in-person design and review sessions.
How to ApplySend a note to [redacted] with anything that helps us understand how you think. The strongest applications include a system you have built or owned end-to-end, a design decision you made and would defend, or a technical bet you took and what you learned. We read everything.
