Junior Finance AI Software Engineer

Accelerate your fintech AI career with our real estate investment client that has a perfect track record on investor capital, and get real ownership while you're at it.

**The Setup**

- Established real estate credit/lending business: niche, not well-known outside the industry, quietly having incredible success

- Riding a massive demographic wave (skilled nursing, senior housing; boomers aren't going anywhere)

- Head of engineering wants to move past one-off AWS Bedrock tinkering into real enterprise AI, and you'd ship the first one

- Underwriting and pre-screening still run on manual work today: highest-impact bottlenecks in the business, handed to you

- 7-person team, flat structure, no formal code review (engineers review each other), no red tape

**What You'd Own**

- A first project tied to a real underwriting/pre-screening bottleneck: visible impact fast, not a 6-month roadmap

- AI-in-the-loop workflows: tight context, predictable costs, clean handoffs (they already learned RAG isn't a magic wand, so you won't relearn that lesson the hard way)

- Bedrock integrations, plus whatever tool actually fits; nobody's married to one stack

- Dashboards/internal tools with the data team, mostly in React

- Full ownership of a problem end to end by year one, not just "the engineering part" of someone else's spec

**Why This One's Different**

- The data is messy: real-world, real-money messy, full of caveats nobody wrote down

- The people who need your help aren't engineers. They're business folks with a vague ask who don't know what they actually need built

- The job: sit with that ambiguity, ask the question nobody else asks, dig through data that isn't pre-cleaned, land on what's actually right, not just what was literally requested

- You'll pick up real credit/lending fluency on the job (what a loan is, how repayment priority works). Nobody expects you to know it walking in, they expect you to want to learn it fast

- Their approach to solving problems gets rebuilt from scratch every ~4 months as the tools and data mature. Priorities you start with in month 1 might look nothing like month 4

**Who Fits**

- About 1 to 3 years out of school

- One real project (professional or personal) where AI was actually in the workflow, and you hit the ugly production stuff: context blowing up, costs creeping, handoffs breaking

- No degree requirement, no name-brand pedigree, no "years of architecture experience"

- A thought provoking side project beats a course list every time

**Growth**

- Advancement tracks impact, not tenure. No waiting a year and a half for a title bump

- Team's most effective engineer built his reputation on curiosity and calculated risk-taking, not pedigree

**Stack**

Python, AWS (incl. Bedrock), Snowflake, Airflow, dbt, React. Fluency not required day one, fast learning is

**Logistics**

- Midtown Manhattan, 5 days/week in office

- Base salary + real bonus upside once you've proven out