Lead AI Architect - Data Platform

<div class="content-intro"><p>Banyan Software is the best permanent home for software businesses that serve specialized industries, their employees, and their customers. With a buy-grow-and-hold-for-life approach and a permanent capital base, Banyan acquires and grows companies worldwide, honoring founder legacies and helping portfolio companies modernize through shared AI expertise and operational discipline. Founded in 2016, Banyan operates more than 120 portfolio companies across North and South America, Europe, and APAC, and has appeared on the Inc. 5000 list for six consecutive years. The Banyan Software Foundation, endowed with $100 million in Banyan stock, leverages technology to build a greener and more equitable world.</p></div><h1>The Role</h1> <p>We are hiring our first Lead AI Architect to design, build, and run the centralized data and AI platform that powers Banyan. The cornerstone of the work is a centralized data model, a Banyan data lake, that consolidates the data we already have and unlocks safe AI experimentation for our internal functional teams: Finance, M&amp;A, Business Development, and others.</p> <p>This role is modeled on a classical Enterprise Architect role and adapted for an AI-first world. You will set the target-state architecture, choose the patterns we standardize on, write code yourself, and own the result. You will partner with our embedded business analysts and the leaders of each functional vertical so the platform stays grounded in real use cases, not abstractions.</p> <p>This is a player-coach role. You start hands-on. Over time you build and lead a small platform team.</p> <h1>What You Will Build</h1> <ul> <li><strong>A centralized data lake on AWS: </strong>the single, governed source of truth for internal Banyan data. Designed for analytics, ML, and AI experimentation.</li> <li><strong>A reference architecture for AI: </strong>patterns and building blocks our verticals use to go from idea to safe production: model selection, RAG, evaluation, observability, cost controls.</li> <li><strong>Guardrails that make experimentation safe: </strong>data classification, access tiering, sandboxes, audit logging, and clear rules for what can be sent where.</li> <li><strong>A platform team and operating model: </strong>the team, standards, and rituals that keep the platform reliable and improving as adoption grows.</li> </ul> <h1>What You Will Do</h1> <ul> <li>Define and own the target-state data and AI architecture for Banyan. Make the trade-offs explicit and write them down.</li> <li>Design and build the AWS-centric data lake. Set the patterns for storage, cataloging, query, ingestion, modeling, and lineage. Likely stack: S3, Glue, Lake Formation, Athena, Redshift, Iceberg or si