Optimization Systems Engineer, Lead

UP.Labs

About The Position

Incubated inside UP.Labs, we're building scheduling intelligence for manufacturers, a vertical AI company focused on the gap between legacy ERP/MRP systems and modern production reality. The current production scheduling tools manufacturers rely on (AS400 green screens, rigid APS solvers, and endless spreadsheet workarounds) weren't built for the daily volatility these teams face. We are building a new system from scratch, with no legacy codebase or inherited architecture, in a 0-to-1 phase. The role is for an Optimization Systems Engineer to own the integration architecture between scheduling and optimization algorithms and the product platform. This person will determine which solvers, heuristics, and optimization approaches to use, how to configure them, and most importantly, how to architect the systems that make them work inside a real product that real planners depend on. This is a lead role with a clear path to technical leadership, offering significant influence over architecture, technical roadmap, and the product.

Requirements

You've built optimization software and you've shipped it—not configured it, not managed the team that built it. You've written the code, debugged the edge cases, and dealt with the fallout when the solver returned garbage because the input data was wrong
You've seen the inside of ERP/MRP/APS systems and have strong opinions about what's broken. You've felt the frustration of watching a planner ignore an optimization output because it didn't account for something obvious. You want to fix that
You're a builder, not an operator. A blank repo and open architecture excites you more than maintaining an existing system
You're an AI-native developer. You use tools like Cursor, think in terms of spec-driven development, and leverage AI to move faster without sacrificing quality—this is how we work
You default to simplicity. You know the difference between elegant and over-engineered. When a heuristic solves 90% of the problem, you don't reach for a MIP solver
Strong software engineering fundamentals—you can architect systems, not just write algorithms. API design, state management, testing strategies, deployment patterns are not foreign concepts
Fluency in optimization paradigms: linear/mixed-integer programming, constraint programming, heuristics, metaheuristics—and knowing when to use what
Experience with production scheduling, supply chain planning, or adjacent domains (transportation, logistics, energy)
Comfortable working in Python
Familiar with modern data infrastructure (Databricks, Delta Lake, or similar) at a level that makes you a good partner to the data team

Nice To Haves

Manufacturing-specific experience is a plus but not required
Experience with C/C++ for performance-critical components is a strong plus

Responsibilities

Design how scheduling algorithms, constraint solvers, and optimization tools plug into our platform
Own the interfaces, data contracts, and execution patterns that connect algorithmic intelligence to our product
Take optimization approaches (MIP solvers, heuristics, constraint programming) and architect production-grade systems around them
Own APIs, orchestration, state management, and rollback patterns
Bring deep knowledge of how MRP runs work, where APS tools break down, and why planners abandon optimization outputs
Translate that knowledge into every design decision
Evaluate and select solvers, libraries, and frameworks
Know the tradeoffs between commercial solvers (Gurobi, CPLEX) and open-source alternatives and recommend based on our constraints, not vendor marketing
Architect the software layer that handles schedule generation, what-if analysis, constraint evaluation, and recommendation delivery
Make it fast, observable, and trustworthy
Work directly with our data science and data platform teams to define what data you need, in what shape, at what latency
Translate fluently between the algorithmic world and the engineering world