Field Software Engineer, Capture Operations

In addition to our core business, Workstream is home to a small, fast-moving Robotics team focused on advancing the next generation of artificial intelligence and robotics. We partner with a leading AI and robotics company to produce high-quality, large-scale egocentric (first-person) video datasets that help train advanced AI models.

Here are our Company Values and we’d love to have individuals who can be a culture add to our team!

What We Offer

• Pay wired directly to your bank account
• $50 Monthly Stipend
• Annual Bonus
• Flexible PTO
• Enjoy company time off during some of the major PH Holidays
• Career fast track for performers and internal mobility
• Learn from a team of Stanford, MIT, Google and Harvard alumni

We are looking for a Software Engineer to join our growing Robotics Data Collections team. This role will keep the footage flowing, from a stuck upload at a capture site to the dashboard screen the operations team reads every morning.

Your users are real people a few metres away: the operators and contributors recording every day. When something breaks, it's their hours, their pay, and their patience on the line, and this role exists to care about that deeply. You'll see your users face to face at capture sites, hear them in the support channels, and build with them in mind.

Every day, capture sites record hours of footage that must land safely in the cloud. When it doesn't (a stuck upload, an unreadable card, an app refusing to finish, hours that don't add up), you are the engineer who finds out why, unblocks it without losing data, and then improves the software so it explains itself next time. The work spans every surface the operation touches: a desktop app site staff use to offload and upload footage, a phone app contributors record with, web dashboards and reports, and the physical devices themselves. You'll build on some of these surfaces and debug on all of them.

You are also the engineering voice site staff hear every day. Our operations run on Slack, across time zones. Operators post a screenshot of an error and ask what it means and what to press next. The core engineering team wakes up hours later and reads what you wrote to understand what happened in the field. This role lives in that conversation, translating between confused screenshots and root causes, in both directions.

This is an AI-first job. The operation is scaling faster than any team could answer tickets one at a time, so we don't. You'll use AI agents as a daily multiplier: to write diagnostics while you're still reading the error, to investigate many stuck uploads at once, to draft the guide while you fix the cause.

What it's like

• Small, fast-moving team; your work shows up in the operation the same week.
• A mix of support conversations, focused software days, and field days at capture sites.
• AI agents are a first-class part of how the whole team works. You'll work directly with the

Day In The Life

Product engineering (40%)

• Build features in the desktop and mobile apps. Ship improvements to the desktop app site staff run every day (offloading cards, uploading, seeing what landed) and the phone app contributors record with: better flows, clearer states, fewer ways to get stuck.
• Build the dashboards and reports the operation runs on. Hours and progress reports, upload accounting, status screens: make the numbers trustworthy and the words on screen say what's actually happening.
• Write software every week, with AI as your leverage, across the desktop app (Electron), the phone app (React Native), web dashboards (React/TypeScript), and Python services and tooling. You'll routinely have AI agents write and iterate on diagnostics, tests, and features with you.

Field support & debugging (35%)

• Run the daily support loop. Be the first responder in our support channels when site staff hit upload, app, or device problems; answer in plain language, unblock them fast, and follow through until it's resolved, not just explained.
• Diagnose and unblock the upload path. When footage is stuck between a capture card and the cloud, trace where and why (device, card, app, network, cloud), get it moving again safely, and make sure nothing is lost or double-counted along the way.

Scaling the operation (15%)

• Solve classes of problems, not instances. Every question the ops team has to ask an engineer twice is a gap; turn recurring ones into self-serve screens, clearer error messages, AI-powered diagnostics, and guides operators reuse. Your goal is that the operation can grow without support growing with it.

Devices & hardware (10%)

• Be hands-on with devices when the problem is physical. Use AI-written tests to check why a camera or battery isn't behaving, open a device to inspect it, and swap parts like cameras and buttons. No soldering. You'll also help put beta hardware in operators' hands and report what breaks.

Who You Are

• Deep care for the user. The operators and contributors using what you build are people you'll know by name; their lost hour bothers you until it's found, and "works on my machine" isn't done.
• Flexible, fast, and scrappy. You'd rather ship a working fix today and improve it tomorrow than design the perfect one next week; you're comfortable when this month's job looks different from last month's.
• A willing debugger and problem-solver. You can follow data through a pipeline with many stages, form hypotheses, write quick tests (with AI's help), and follow the evidence rather than guessing, wherever the problem lives.
• AI-leverage mindset. You already use AI tools well and they multiply you; when problems repeat, your instinct is to build something that solves the whole class rather than working through them one by one. You treat the challenges of a scaling team and operation as the fun part, not the fine print.
• Strong communicator, concretely. With site operators and contributors, you answer "what does this error mean?" with what it means and what to press right now, in plain words, and you write step-by-step guides a non-technical person can follow on the first try. With the operations team and core engineers, you write down what you saw and tried, pick up a thread someone left overnight, ask good questions, and keep a problem moving across time zones until it's solved.
• A collaborator across teams. You stay patient on the fifth "it's still not working," and people finish a conversation with you knowing exactly what to do next.
• Care for data integrity. In our world, a lost or double-counted recording is someone's unpaid hours; you treat that seriously.
• Self-directed. You'll often be the only engineer on site, supported by a remote team in other time zones.

Preferred Qualifications

• Solid software fundamentals: working Python, comfort in Linux (services, logs, networking), and the ability to build app UIs
• Strength in at least one of web (React/TypeScript), desktop (Electron), or mobile (React Native), and the willingness to become productive across all three and debug anything
• Daily, effective use of AI coding tools and agents
• Clear written English for remote, asynchronous work
• Located in Metro Manila, and able to travel to capture locations regularly

Nice To Have

• Experience shipping Electron or React Native apps to real users
• Experience in a support-adjacent engineering role: field engineering, solutions engineering, IT for operations teams
• Hardware tinkering: Raspberry Pi and single-board computers, cameras, taking devices apart; 3D printing for jigs and mounts
• Familiarity with cloud storage, upload pipelines, SQL/Postgres, or fleet/IoT management