Staff Software Engineer

The Problem You'll Solve

Industrial control panels are built from hundreds of components: contactors, terminal blocks, PLCs, switches, and more. Each component has rich, structured data — terminal semantics, electrical ratings, spatial information, certifications — that our design software depends on to generate accurate, manufacturable panels.

Very little of this data is accessible in machine-readable form. It is buried in complex, inconsistent sources like PDFs, user manuals, datasheets, STEP files, and diagrams. Structuring this data manually does not scale.

Your job is to solve this problem at the root. You will design and build systems that extract, validate, and serve trustworthy information from these challenging data sources.

What You'll Do

• Platform Ownership: Own the technical roadmap and expansion of our proprietary data model for industrial devices.
• Extraction Pipeline Design: Architect and build extraction pipelines that pull structured data from PDFs, datasheets, user manuals, STEP files, images, and manufacturer APIs. The right solution might involve purpose-built parsers and validators, off-the-shelf tools like OCR libraries, ML-based approaches, or some combination — you'll determine the best fit.
• Continuous Improvement: Extraction is not the end of the story. We iterate on and refine component data, gaining trust in it over time. You will add confidence scoring, human-in-the-loop review, and rollout strategies. You’ll expose device data as a landscape with varying degrees of accuracy, completeness, and effectiveness.
• Cross-Functional Leadership: Collaborate closely with electrical and process engineers to identify data needs and translate them into software solutions. Serve as the technical leader for this domain and a mentor for other engineers.
• Strategic Growth: Proactively identify and pursue new opportunities to improve the business through the application of technology. We are a software-enabled hardware company, and when an unexpected opportunity presents itself we take it.

Our Stack

We use Deno, Postgres, SQL Server, HTMX, Docker, Terraform, and AWS. As the owner of this domain, you have the latitude to adopt whatever tools, frameworks, models, or services best fit the problem.

Who You Are

There are 4 values that all Podium employees are expected to demonstrate (albeit imperfectly):

• Win for the customer. We sit on the same side of the table as our customers - their wins are our wins. We’re building the vendor we always wished we had.
• Default to transparency. We believe in honesty, transparency, and kindness. It’s about how you get things done and not just what gets done.
• Speed is a feature. We move fast and take ownership. Don’t let perfect be the enemy of good.
• Win and lose as a team. Individuals drive inputs, teams drive outputs. We celebrate success together but aren’t afraid to fail and iterate. We own outcomes together and don’t deflect blame.

Required Qualifications

• Experience: 8+ years of software engineering experience, with a record of owning and shipping production-grade systems end-to-end. 4+ years of experience in a startup environment.
• Domain Expertise: Deep experience tackling the problem of extracting and structuring information from messy, inconsistent real-world data sources.
• Pragmatism: You know when to build versus buy, when a simple heuristic beats a model, and when it's worth the complexity of something more sophisticated.
• Location: Commitment to a highly collaborative, on-site work environment in our Brooklyn, NY office 5 days/week.

Bonus Qualifications

• Experience in applied ML, computer vision, NLP, or document understanding (OCR, layout parsing, etc.)
• Experience with the Deno or Node.js runtime environment
• Familiarity with industrial controls or electrical design