Senior Python Backend Engineer, Financial Systems

AI for Makers
We're Makers. In 2025, we set out to build something most teams only talk about. We embedded AI so deeply into our development workflows that we were over 10x more productive. Being on the bleeding edge isn't a strategy here. We hack on AI tools because we can't help it. It's how we're wired.

We don't talk about the future of engineering. We're living it. 

Logistics
This is a fully remote based role, however candidates must be located in Argentina or Brazil. Note: this is a long-term Independent Contractor position.

Please submit your resume in English.

Our tools & stack
- Python, MySQL, AWS
- Claude, Codex, Cursor, Copilot, LangChain, LangGraph, LangFuse, OpenRouter, FastMCP

What you'll do
 - Push production AI agents into the platform’s most complex flows, and build the Python services and MCP servers that orchestrate them: tool layer, context handling, model routing through OpenRouter
 - Build the integration backbone the platform runs on: payments, identity, e-signature, real time data, at the latency live decisions demand
 - Build out the rules engines and configuration runtime that turn the platform into a product factory, with customers shipping entirely new products on top of us without us touching the code
- Migrate whole companies onto the platform: importer pipelines, batch reconciliation services, zero data loss
- Build and scale the multi-tenant batch systems that move real money every night: settlement runs, document pipelines, async workers, where every customer’s data has to stay correct under load
- Design the public APIs the entire customer ecosystem runs on, with AI agents as first-class consumers and the contracts that let new endpoints ship at speed
- Build the financial systems behind every dollar that moves through the platform: billing, payouts, commissions, refunds, every cycle

What you'll bring
- Bachelor’s Degree or higher in Computer Science or a related field
- 5+ years of experience in Python web development, particularly with financial systems or modules
- 3+ years with SQL databases (e.g., MySQL, PostgreSQL)
- 3+ years of SaaS product development, including cloud based solutions
- 3+ years of hands on experience with Django or another Python framework
- 2+ years of proven experience building and consuming APIs for financial transactions
- Daily, deliberate use of agentic AI coding tools such as: Claude, Cursor, Codex
- Proven ability to direct AI agents on multi file, long horizon tasks
- Advanced mathematical skills in financial modeling, algorithms, and data processing
- Hands on experience with frontend frameworks like Vue.js or React
- Professional experience working with Docker for containerization and AWS
- Excellent verbal and written communication skills

Bonus points
- 2+ years of experience working at a scaling startup
- Experience with spec driven development: using AI to generate tests, scaffolding or architecture from requirements
- Hands on experience building production AI agents or multi-step LLM pipelines using LangChain, LangGraph, or equivalent orchestration frameworks
- Proven use of Agent Teams in Claude Code, multi agent orchestration via Codex, or similar parallel agentic workflows
- Experience working in the fintech or financial services vertical

We used tokens
This isn’t your typical AI generated job description

What did we do different?
Other companies: Basic AI
Us: Claude Cowork (Opus 4.7)

Why?
The same reason you’re still reading the job description

Here was our first prompt
I need to use Cowork to update a job description for a Senior Python Backend Engineer role. The current job description is in a folder on my desktop called "Job Description"

I want Cowork to pull from three Slack channels: #eng-general-int, #product-general-int and #ai-for-makers - plus any design docs from the last three months, to understand what we've actually been shipping. Then update the job description based on what it learns

Final output should be a Google Doc. Help me write the Cowork prompt

[Our recruiter wrote this prompt, not an engineer]