Vice President, Platform

Glendale, CA — on-site, 5 days a week

Full-time Compensation: $175,000-$250,000

Bambee started as a subscription service delivering HR support to small businesses across the United States, and grew into a leading provider of direct HR guidance for small and midsized companies.

For the last two years we have concentrated nearly all of our R&D capacity — and a disproportionate amount of founder and executive attention — on Bionic, our proprietary AI architecture. That work defined a new category: Cloud Workforce Operations, the continuous management of the systems that determine how a workforce operates, including its people, policies, performance, communication, compliance, and organizational health.

Hexagon is the platform at the center of it. Instead of waiting for a company to ask for help, Hexagon observes what is happening, understands what should be true, identifies the gaps, decides what happens next, and coordinates or executes the work to close them. Every completed operation makes the system more capable.

Every major product and operating decision Bambee makes over the next five years will be made cheaper, faster, or more expensive by architectural choices made in the next twelve months. We are looking for the person who owns those choices.

The Vice President, Platform owns Bambee's Intelligence & Leverage System: the shared technical and intelligence architecture through which data, models, reasoning, memory, agents, orchestration, and reusable platform capabilities make every Bambee product and internal operating system more capable.

You own what Bambee's systems can know, reason about, remember, execute, and reuse. You decide what gets built once and inherited everywhere, where system boundaries belong, what must stay stable, what should be deprecated, and where temporary speed creates unacceptable long-term cost.

Our operating principle is Outcome → Intelligence → System → Human Judgment → Headcount. The sequence is intentional, and you are one of its primary architects.

You will be judged by whether every new product, agent, workflow, and operating system inherits more capability than the one before it.

1. Intelligence architecture. The architecture connecting organizational data, persistent context, models, reasoning, memory, evaluation, learning, and execution. You will define how the system holds company context, retrieves memory, invokes models, evaluates outputs, and improves its next decision. The objective is not more models — it is one coherent intelligence system that products inherit instead of rebuilding.

2. Primitives, reuse, and compounding. Which capabilities get built once and inherited everywhere: identity, permissions, organizational state, memory, orchestration, agents, execution, evaluation, auditability, documents, integrations. You also own the tooling, testing, evaluation, and deployment architecture that lets teams move independently without recreating foundational work. The measure is whether the next capability is easier to build than the last one, because the platform already contains more of what it needs.

3. Autonomy and orchestration. Hexagon already observes, assesses, diagnoses, recommends, coordinates, and performs selected categories of work on its own. That scope will widen. You will architect how agents are orchestrated, how work routes between machines and humans, how confidence and risk determine autonomy, how exceptions are handled, and how outcomes are verified — more operational capability, with real control and observability.

4. Data, extensibility, reliability, and security. The core data model, system-of-record architecture, APIs, integrations, versioning, multi-tenant boundaries, reliability, security, access control, observability, infrastructure cost, and backward compatibility. You will define how organizations, employees, policies, assessments, recommendations, actions, and outcomes are represented, and build the integration surface connecting Hexagon to the systems customers already use. Reliability, security, and infrastructure economics are architectural inputs, not downstream concerns.

5. Organizational leverage. The platform should make Bambee itself more capable, not only its products. You will work across Finance, Product, Growth, Operations, Customer, and Frontier to find repeated cognition, fragmented information, duplicated logic, and manual coordination that should become shared system capability. When several teams solve the same problem, ask whether the platform should solve it once. When people are acting as middleware between systems, ask whether that coordination can be encoded. When a function needs linear headcount to handle rising complexity, ask whether the platform can change the economics of the work.

We are not optimizing for a conventional platform-management résumé. The strongest candidate may have been a CTO, Chief Architect, Head of AI Platform, VP Engineering, or a technical founder:

• Technical founder or former CTO who built foundational architecture, then had to connect technical decisions to products, customers, and economics
• AI or ML platform architect who moved past isolated modeling into production systems involving context, retrieval, reasoning, agents, orchestration, and evaluation
• Data or platform leader who grew from shared services into systems that make data actionable through models, reasoning, and automation
• Distributed-systems architect with strong product judgment
• Technical product or platform leader who owned shared architecture while staying close to what the business needed it to enable

We care more about the systems you have personally shaped than the progression of your titles.

Architecture depth. You reason deeply about data, distributed systems, APIs, AI systems, reliability, security, and cost. You can descend into the architecture rather than managing it from abstraction, and you know when an attractive abstraction is masking a weak boundary. You have made consequential architecture decisions and built judgment from the ones that worked and the ones that did not.

Intelligence-systems thinking. You understand that AI infrastructure is larger than models, and can reason about how context, memory, retrieval, agents, orchestration, evaluation, and feedback loops form a single coherent system. You can tell product-specific intelligence from shared infrastructure, and you know the quality of an intelligent system depends on what it knows, remembers, and verifies — not only on what it can generate.

Leverage instinct. You would rather delete two redundant systems than build a third one to coordinate them. You get suspicious when several teams solve the same problem independently. You think about architecture in economic terms: marginal cost, coordination cost, reuse, compounding capability. You will say no when short-term speed creates structural cost, and you will not use architecture as an excuse for delay.

We take on some of the hardest problems in workforce operations because solving them unlocks disproportionately ambitious outcomes. We are building it for ordinary companies and the people who spend their lives inside them.

If that sounds like the kind of company — and the kind of system — you want to help build, we should talk.