Software Engineer

Location: Remote (US). Chattanooga, TN or nearby is a plus.

Type: Full-time

Team Size: Small, high-impact product team

About Legacybox

We help families preserve the things that only exist once.

From VHS tapes and film reels to printed photos, negatives, slides, DVDs, camcorder cassettes, memory cards, and phones, we digitize the media that matters most. Families can then revisit and share those moments for generations to come.

But we believe the future of Legacybox is much larger than digitization alone. We are building a long-term family memory platform designed to help families preserve, organize, explore, and share their history together.

We are not in the digitization business. We are in the irreplaceable business.

How we build

The old split, where engineers code, designers design and product managers write specs, is breaking down, and we're leaning into it. Agents write a lot of our code. Prototypes take an afternoon. We're building internal tools so the whole company works this way, not just the product team.

So here's who we're looking for: you've done one of these jobs (engineering, product or design) well, and you loved doing it. Here you'll use AI to cover the other two, next to teammates who are strong where you aren't. We care as much about how you check the AI's work as how fast you get it done.

What we're building

Digitizing is where a family starts with us. It isn't where we want to stop.

Most families have decades of their history in boxes, on formats almost nobody can play anymore. Getting it digitized solves the first problem. The second is that a digital archive can sit just as untouched as the box it came from.

We think a family archive should give people a reason to come back. Hours of tape hold things nobody remembers recording: a grandparent's voice in the background, the first time two cousins met, a house that's gone now. Most of it is never watched closely enough to find. We want to surface those moments, and help families see things in their own history that weren't obvious before.

We also want the archive to be the place a family's memory collects. The recordings and photos are only part of it. The rest lives in people's heads: who's in the picture, where it was taken, what happened after. We want families to add what they know while the people who know it are still around, so the archive gets more complete over time, not less.

The role

You'll work across all of our software, not one corner of it. Where you start depends on you and on what matters most when you join. Some people lean toward what customers see, some toward the systems our own team runs every day. Either way, you'll be productive across all of it.

What you'll work on

• The Platform: Python services on Google Cloud behind our web, iOS and TV apps: orders, memberships, media, sign-in and notifications.
• The customer web app, where families watch, organize and share what we digitized.
• Our mobile and TV apps
• Shopify: our storefronts, our own Shopify app, and the checkout and admin extensions that tie the store to the Platform.
• The Digifactory: the software our team in Chattanooga uses to digitize every tape, reel and photo. That's the floor system, capture and encoding at each station, and the tools that check quality.
• Media at scale: petabytes of family video, audio and photos across Backblaze, AWS and Cloudflare.
• Data: the warehouse and pipelines that bring store, subscription, marketing and Digifactory data into one place.
• AI in the product and on the floor: Gemini on Vertex AI to understand media today, with transcripts and photo enhancement next. All of it runs under clear rules for how we treat customer media.
• New ways in: member uploads and capture in the app (voice notes, stories, photos). Search across a family's archive, and 3D scenes and models built from photos and video.

Real work on our list

• Help a family find the five minutes that matter in forty hours of tape.
• Make the day a customer's order lands in their library feel like opening the box, not downloading a folder.
• Move the web app's library and sign-in off our legacy API and onto the Platform with shadow reads, feature flags and staged rollouts. No big-bang cutover.
• Retire older AWS serverless systems one piece at a time: (1) write to both systems, (2) backfill millions of rows, (3) hit the off switch.
• Build checkout features with Shopify checkout extensions, Shopify Functions and Recharge subscriptions, backed by Platform APIs.
• Turn our station capture app into a real product: signed installs, staged updates across the floor, telemetry.
• Send customers progress updates as their order moves through the Digifactory, in the app, by push and in email digests.
• Add transcripts and search to members' libraries, with AI running in the cloud and on our own hardware.
• Let members upload their own photos and video into their library, stored and processed like everything we digitize.

Your first 90 days

• Week 1: something small in production, and a day on the Digifactory floor (in person or on video).
• Day 30: you own one piece of one system, from the first pull request to the person using it.
• Day 90: you've shipped in at least two of our systems and you own a lane. You're also the second person who can keep our data pipelines running.

Our stack

• Platform: Python 3.13, FastAPI and Postgres 16 on Google Cloud: Cloud Run, Cloud SQL, Pub/Sub, Cloud Scheduler and Cloud Tasks, Apigee, Secret Manager. All of it in Terraform, shipped with GitHub Actions.
• Web: Next.js 16, React 19, TypeScript, Tailwind, Drizzle, Vercel
• iOS and TV: Swift and SwiftUI, and React Native for TV
• Shopify: Liquid themes and our own app (React Router and Polaris) with admin, customer-account and checkout UI extensions. Shopify Functions, the Admin GraphQL API, webhooks, metafields and Recharge.
• Digifactory: Laravel 13 (PHP 8.5) with React. Python and FFmpeg with GPU encoding at the stations.
• Media: Backblaze B2, AWS S3 and Glacier, Cloudflare Workers
• Identity, messages and analytics: Auth0, Knock, PostHog for product analytics and feature flags
• Data: BigQuery and Dataform, plus our own internal analytics site
• AI: Gemini on Vertex AI, local models on our own hardware, and the 3D and generative media tools we're adding
• How we write it: Claude Code, Cursor and other agents. You review what they write before it ships.

What we're looking for

• 2+ years shipping production software that real people use. Show us something.
• Strong in at least one of Python, TypeScript/React or PHP/Laravel, and willing to work in all three.
• You've deployed something to the cloud and debugged it in production.
• AI is how you work every day: agents, not autocomplete. You can tell us the last thing an agent got wrong for you and how you caught it.
• You talk to the people who use what you build, and you can write a one-page brief before you write the code.
• You measure what you ship.

A plus

• Google Cloud, Terraform, or Postgres at scale.
• Shopify apps, Shopify Functions or checkout extensions.
• Media at scale: video, audio, photos, FFmpeg.
• Computer vision, speech-to-text, search, or 3D (photogrammetry, Gaussian splatting or similar).
• A data pipeline or SQL model someone relied on.
• Software that runs on machines you don't control: kiosks, stations, a mixed Windows and macOS fleet.
• Moving off a legacy system without breaking it.
• Swift or React Native.
• Consumer or subscription products.
• You're in or near Chattanooga, or happy to visit regularly.

How we work

• We put structure behind our goals. Each quarter has a few clear goals, each with an owner and a number to move, broken into outcomes we can land in weeks. Owners come from across the team, and that can be you.
• We ship small and often. Every launch comes with its rollout: who needs to know, what support will say, and how to stop it.
• AI is how we work. Agents do a lot of the building. People own the judgment: what to build, whether it's right, and how it lands.
• We're evolving on purpose. How a small team works with AI is changing fast, and so are we. We try new ways of working, keep what helps, and drop what doesn't.
• We work like professionals. Plans and decisions are written down where anyone can find them. We collaborate across roles, say the hard thing early, ask for help, and follow through.
• Remote, with time together. We're spread across the US and get together in Chattanooga a few times a year, including time at our digitization facility.

Where this goes

This role starts above entry level, with a path to Senior, Staff and Principal Engineer. You move up by what you own and ship, not by years served.

Interview process

1. Intro call with Adam (30 min). Your work, our work, and how you use AI today.
2. Work sample (a few hours, on your time). A small piece of real work in your craft, plus one piece outside it. For this role that's usually a small feature plus the one-page brief behind it. Use whatever AI tools you use every day. We'll ask how you checked the output.
3. Deep dive (60 min). Show us what you built and why.
4. Final round. Meet the team. Depending on where you are, that's in Chattanooga (we cover travel) or on video, and sometimes a paid trial day of real work.

About 5-8 hours over 2-3 weeks.

How to apply

We’ll ask you to show us something you built with AI in the last three months that's outside your main craft.

Compensation & benefits

• Competitive salary based on the scope of responsibilities you’re qualified to be assigned.
• Health insurance: medical, dental and vision, with 100% of the employee premium covered
• 401(k) matching
• PTO, and we actually want you to use it
• Remote work from anywhere in the US
• Professional development: conferences, courses, books
• MacBook Pro and the gear you need
• Quarterly team offsites
• Free Legacybox service for your own family's memories

Legacybox is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees. All employment decisions are based on qualifications, merit, and business needs.