Senior DevOps Engineer - platform Azure & Full Stack

Overview

Senior Engineer — Platform, Azure & Full Stack FidTech Partners — Remote or hybrid — Full-time — Reports to the CTO

About us : FidTech Partners builds software that community foundations and philanthropic organizations use to run their funds, gifts, grants, donor relationships and reporting.

There are three products:

FoundationHub for foundation staff, DonorSphere for their donors, and GranteeSphere, our newest, which ships its first release in October.

When a foundation approves a grant or sends a donor their quarterly statement, that is our software doing it. It is real money and real donor data. Getting the number right matters more here than shipping fast.

Read this part before the bullet points

This role is not a pure SRE position, and it is not a pure application-developer position. It is both: you write features in C# and React alongside the rest of the team, and you own the platform those features run on.

If you only want to do one of those two things, this is the wrong posting, and we would rather you knew that now.In exchange, you get something small companies can offer and large ones cannot: you will not have to lobby three committees to fix something that is obviously broken. You will fix it.

What you would own

The Azure estate. Multiple environments from development through to production: Function Apps, App Service and IIS, Azure SQL and SQL Server on VMs, Key Vault, App Configuration, virtual networks and private endpoints, managed identity end to end.Delivery pipelines across roughly forty repositories. GitHub Actions with OIDC federation, infrastructure as code in Bicep.

The database estate. SQL Server, a separate database per client, replication, backup, restore and disaster recovery. This is not a DBA role, but you cannot be squeamish about databases and do this job well.

Observability and incidents. Application Insights, alerting, and the internal operations dashboard we built to watch all of it. When production misbehaves, you are one of the people who finds out why.

Security posture. Entra ID, conditional access, WAF and secrets hygiene.

And the products themselves. Not as a footnote: you will write a great deal of application code.

GranteeSphere, launching in October, is our newest product and the most modern thing we have built. It includes a drag-and-drop grant application builder, a review-workflow designer, and AI-assisted scoring of incoming applications. The stack is current: dot.NET Core 10 on the back end, and React 18 with TypeScript in strict mode on the front end, with Vite, Redux Toolkit with RTK Query, Tailwind, Zod-validated forms, Storybook, Vitest and Playwright. It was designed properly, from scratch, this year. If you have spent the last three years fighting somebody’s decade-old stack, this is the antidote.

FoundationHub is the opposite end: a large Classic ASP application, roughly twenty years old, carrying an enormous amount of hard-won financial logic. We are migrating it onto React and .NET screen by screen while foundations use it every day. This is difficult, careful, interesting work, and it is nothing like a greenfield build.

Most engineers only ever get to see one of those two worlds. Here you get both, usually in the same week.What we are looking for

The core:

Enough years to have formed opinions and to have been wrong about a few of them. Call it seven or more, though we will judge the work rather than the number.

Strong C# / .NET. Our newest service is on dot.NET Core 10; our oldest is rather further back. You have designed REST APIs rather than just inherited them, and you have the judgment to know when a background job beats an endpoint.

Comfortable in React and TypeScript. You do not need to be equally strong on both sides, but you should be able to take a feature all the way to the screen without waiting for someone else. Strict-mode TypeScript, modern React and a tested component library are the standard we hold our new work to.

Azure in production, not in a tutorial. You have deployed it, broken it, debugged it, and written the infrastructure as code afterward so it did not happen twice.

SQL Server you can tune. Execution plans, indexing, and the confidence to open a two-thousand-line stored procedure without flinching.

You have owned something operationally. You have run an incident, written the post-mortem, and shipped the fix that stopped the alert recurring.An automation reflex. If you have done it by hand three tim…