Founding Mainframe Engineer

Hypercubic

18d•Hybrid

About The Position

At Hypercubic, we’re building AI systems that can understand, reason about, and rearchitect large-scale legacy applications. We work at the intersection of program analysis, knowledge extraction, and agentic systems, applied to some of the most critical enterprise software in the world. Learn why we're excited about mainframe modernization. About the Role We’re looking for a founding mainframe subject matter expert with hard-earned domain authority. You have spent years working directly inside production mainframe environments, understand the broader mainframe tooling, and have been personally involved in real modernization efforts from technical discovery through execution. You will work closely with the CEO and CTO to shape Hypercubic’s product and technical direction.

Requirements

You care deeply about your work and the product.
You think in systems and feedback loops, not just features.
You have hands-on experience reading, debugging, and modifying large COBOL codebases.
You want to help build world-class AI for mainframe modernization.
You are comfortable working at the edge of what's possible, applying legacy wisdom to the next generation of software engineering.
You have a bias for action and a track record of working in fast-paced environments with urgency.
You can adapt your deep mainframe knowledge to modern engineering and AI paradigms.

Nice To Haves

Have 10+ years working directly with production mainframe systems.
Direct involvement in at least one real modernization, migration, or large-scale refactoring effort.
Experience leading technical discussions with enterprise customers.
Comfort owning outcomes and having the technical judgment to lead through ambiguity.

Responsibilities

Be the Authority: Act as a hands-on technical authority on mainframe systems, including COBOL, JCL, RPG, CICS, and the surrounding legacy ecosystem.
Bridge the Gap: Act as the technical link between 40-year-old system architectures and modern, agentic AI paradigms to ensure the output is industrially sound. You also have the ability to map JCL-driven batch step dependencies and CICS Task Control blocks to modern asynchronous or microservice-based workflows.
Direct Product Roadmap: Pressure-test Hypercubic’s product against real-world constraints such as batch processing, data dependencies, operational risk, testing complexity, and regulatory requirements.