Senior Software Engineer, Integrations

Element451

Element451 is building the AI-powered platform reshaping how colleges and universities recruit, enroll, and support students — and integrations are the connective tissue that makes it work. You’ll own the data pipelines, third-party API integrations, and distributed workflows that move student data reliably at scale, with AI built into how you work. It’s a chance to own meaningful systems end to end, treat AI as real craft, and set the standard others in this space measure themselves against.

The Role

Remote in Serbia, you’ll build integration software with full end-to-end ownership — from shaping a pipeline in planning to keeping it stable in production. This is an explicitly AI-forward role: AI tooling is a professional discipline applied across every phase of the work, not a shortcut for generating code. Senior Engineers operate with autonomy, hold their work to a high bar, and lift the quality of the work around them. We’re upfront about that bar because the engineers who thrive here want exactly that.

What You’ll Own

You own the integrations you ship until they’re live and moving real student data correctly. There’s no separate QA function: “done” means work you’ve verified yourself by running real customer records through the integration — not code that merely merges and runs. Day to day, that looks like:

• Build and own data pipelines and integration services designed for idempotency, safe retries, and observability from the start.

• Treat data integrity as sacred — integrations that silently drop, duplicate, or corrupt records are worse than ones that fail loudly — and catch discrepancies before a client does, pulling in the Integrations Architect on higher-risk changes.

• Drive root causes to resolution across incidents, sync failures, and data discrepancies, rather than patching symptoms or papering over a flaky upstream API.

• Surface production risks early — schema drift, upstream API changes, rate limits, credential expiry, partial-failure edge cases — before they become incidents.

• Generalize fixes that help more than one integration into shared libraries, escalate broad-blast-radius decisions to the relevant Principal Engineer, and drive completed work all the way to release.

On AI: you treat code assistants, test generation, and debugging as core craft, and hold AI-assisted work to the same bar as anything you write by hand — every output reviewed, tested, and owned before merge. Speed never substitutes for judgment, especially in data integrations, where plausible-looking but subtly wrong code corrupts records silently.

How You’ll Show Up

We hire for behavior as much as experience. These are the values we live by and hire for:

• Understand the “why” before the “what.” Grasp the real problem before building, and question a requirement that will quietly break for the next client.

• Own the outcome. It’s done when it’s live, validated against real data, and stable in production — not at merge.

• Take initiative. Move work forward and surface risks while they’re still cheap to fix, without waiting to be told.

• Solve problems together. Share openly and treat feedback as a professional input; the best solutions come from the team.

What You Bring

• 5+ years of software engineering, including senior-level work on data integrations or backend services in a complex, multi-tenant SaaS product.

• Strong, current Python for production data pipelines and integration services.

• Solid MongoDB — data modeling and query performance across flexible, document-oriented schemas at multi-tenant scale (~140 client integrations).

• Distributed systems — async, queue-driven architectures and the realities of partial failure, retries, idempotency, eventual consistency, and backpressure.

• Third-party API integrations (Student Information Systems, Stripe, PayPal, Google, Outlook) — auth, pagination, rate limits, and graceful handling of unreliable upstreams.

• AWS-hosted environments (secrets management, SQS, workflow orchestration), and comfort in an async-first, PR-based team with rotating production-reliability ownership.

• Automated testing across layers, plus sound AI judgment — production experience with OpenAI APIs or comparable LLM platforms is a plus.

• Deep, current fluency with agentic AI development tooling (Claude Code, Codex, or comparable), and excellent written communication that stands on its own.

Our Interview Process

Our process is rigorous and designed to be a real signal. Expect live sessions with our engineers where you’ll demonstrate your work and reasoning in real time, including how you apply AI tooling. We move quickly for the right people.