Software Engineer

About Us

HuxleyIQ is a small team building something genuinely powerful: AI-powered investment due diligence that compresses weeks of analyst work into minutes. We're looking for a programmer who wants to ship real product alongside people who care deeply about the craft.

You don't have to be a 10× developer or a machine-learning PhD. You don't have to love Jira or hustle culture. You do have to love beautiful code, care about users, and be the kind of programmer who learns about the business goals before diving into the implementation.

We built HuxleyIQ on Ruby on Rails because it lets a small team punch far above its weight. Our stack is a majestic monolith: Rails 8, Hotwire (Turbo + Stimulus), PostgreSQL, Solid Queue, and a purpose-built multi-agent AI workflow layer. Everything is multi-tenant, account-scoped, and built to handle real PE and family-office deal flow.

About the Job

HuxleyIQ's core product creates structured, actionable analysis for financial firms. A multi-agent AI pipeline extracts financial data, maps it against a firm's investment thesis, surfaces risks, and drafts IOIs. That pipeline runs in production today, used by real investors on real deals.

As our next programmer, you'll work across the full stack. Some cycles you'll be deep in the AI workflow layer — designing how agents coordinate, how artifacts are stored and surfaced, or how results are structured for downstream use. Other cycles you'll be on the front end, pairing with design to ship polished Hotwire interfaces that investors will rely on daily. You'll also invest time in keeping the codebase healthy: squashing N+1 queries, writing tests that actually test something, and maintaining the multi-tenant account isolation that our customers trust us to get right.

About You

We're a small team with a lot of surface area. No one is going to hand you a ticket queue and tell you exactly what to do each morning. You'll see work that needs doing, take ownership of it, and see it through — from the first commit to production and beyond.

We trust quickly. We hold ourselves accountable. We believe that code is writing — clear thinking produces clear code, and clear code produces happy users.

Some specifics we're looking for:

• Solid Rails experience. You know how Active Record works under the hood, you reach for Hotwire before JavaScript frameworks, and you don't need to look up how strong parameters work.
• You write tests — real ones. Not just happy-path smoke tests, but tests that cover the edge cases where real bugs hide. You've used something like Minitest or RSpec seriously.
• You understand multi-tenancy. Account-scoped data isn't just a pattern to you — it's something you think about automatically when writing queries and designing features.
• Curiosity about AI. You don't have to be an ML engineer, but you should be genuinely interested in how LLMs work and excited to build on top of them. Knowing what a prompt is, isn't enough — we want someone who thinks carefully about how to structure agents, outputs, and coordination.
• An excellent writer of prose and code. The bulk of our work — pull requests, feature proposals, async communication — is written. Being clear in text makes you a better teammate here.
• Comfort with ambiguity. We're early-stage. Priorities shift. Sometimes the right answer is to build a thing; sometimes it's to talk to a user and figure out whether the thing is even right.