AI / Backend Engineer — First Distributors (Houston, TX)
We're a Houston industrial PVF distributor building a platform that turns messy customer RFQ emails, PDFs, and spreadsheets into structured, priced quotations.
The software is live with our own sales team today and we're building toward agents that can run the full quoting workflow end to end.
You'd be the second engineer on the team, working directly with the CEO and with our lead engineer. You'll have unusually high ownership over architecture, product direction, and how we solve difficult technical problems.
What you'll work on
- Document intelligence: extracting structured BOMs and line items from highly inconsistent emails, Excel workbooks, PDFs, tables, and other documents that were never designed to be machine-readable
- Reliable AI pipelines: combining LLMs, deterministic parsing, rules, confidence scoring, validation, retries, and human review rather than treating every problem as a prompt
- Evaluation infrastructure: golden datasets, regression suites, and accuracy metrics we actually trust before shipping a prompt or model change
- Product matching and entity resolution: matching messy industrial descriptions against hundreds of thousands of SKUs
- Workflow automation: tool use, multi-step processes, third-party integrations, retries, exception handling, and knowing when software should hand a problem to a human
- Production backend systems: Python/FastAPI services, PostgreSQL data modeling, background jobs, APIs, authentication, observability, and integrations with external systems
- Full-stack delivery: making changes across the backend and our React/TypeScript web application when necessary to ship a complete solution.
What we're looking for
- 3+ years of production backend engineering in Python; you write real services, not notebooks
- Hands-on experience shipping LLM-backed features to actual users, prompt design, structured output, function calling, handling non-determinism
- Strong SQL and data-modeling instincts
- A rigorous approach to reliability: you measure before you optimize and want to know whether a change actually improved the system
- Good judgment about when not to use an LLM and when deterministic code, rules, or traditional algorithms are the better solution
- You measure before you optimize. If you've built or wanted to build an eval harness because you were tired of guessing whether a change helped, you'll fit in well with our team.
- Comfortable in a startup environment: no PM, no QA, no ticket queue. You'll talk to the sales reps yourself and decide what to build
- A rigorous approach to reliability. You want to measure whether a change actually improved the system rather than relying on a few good examples.
- Ability to take ownership of an ambiguous problem: investigate it, propose an approach, implement it, test it, deploy it, and iterate based on real user feedback
Especially valuable experience
- Document AI, table extraction, complex Excel parsing, PDF parsing, OCR, or multimodal document understanding
- Vector search / retrieval systems at scale
- Building evaluation frameworks for LLM or document-processing systems
- Agent frameworks—or strong opinions about when they're unnecessary
- Domain-heavy B2B software such as distribution, manufacturing, logistics, supply chain, or ERP
- AWS/Azure, Docker, CI
- Front-end experience: React and TypeScript. Not required, but our users work in a web app all day and being able to ship a fix end to end is a real advantage
- Experience working at an early-stage startup or on a small engineering team where you owned major features end to end.
Where is the Role Located
- Houston, TX – near IAH Airport
- Role is on-site with the option to work from home on Fridays
Why this role is different
This isn't an AI demo looking for a use case. The software is already processing real RFQs for a real industrial distributor, and every mistake/win has a real impact.
You'll work on difficult problems involving messy real-world data, ambiguous product descriptions, document understanding, large product catalogs, and AI systems that have to be reliable enough for people to trust with actual high-dollar purchasing and quoting decisions.
You'll also be joining extremely early. The engineering team is small, and you’ll have direct access to leadership and users.
You won't be joining to own one narrow layer of the stack. We want engineers who can understand the business problem, determine the right technical solution, and carry it all the way into production.
As employee number 2, this role comes with significant upside potential and upward mobility as our vision takes shape and the company grows.
Pay: $100,000.00 - $150,000.00 per year
Benefits:
- Dental insurance
- Health insurance
Application Question(s):
- Why do you believe you would be a good fit for our team?
- What interests you in this position?
- This role is in person near the IAH airport in Houston, TX. Will you be able to work from 8am-5pm on-site?
Work Location: In person
