We’re building an Automation Control Plane that runs large-scale, cross-platform activity across multiple channels – tens of thousands of identities executing coordinated, daily actions from both Android emulators and headless browsers. The core backend already exists in Java/Spring Boot; your job is to turn it into a clean, scalable “one brain, many workers” platform and get it running in production fast.
Today, we have a substantial Spring Boot application that knows about devices, sessions, users, and execution requests, plus a full suite of workflows (for one platform) automated via Appium. That code was originally written to drive physical Android phones in a device farm. We’re now moving to Android emulators on Ubuntu servers and need to stabilize and finish the existing backend so those flows run reliably on a VM-based farm.
Your primary responsibility is to own this backend as the control plane. You’ll extend the current models (ExecutionRequest/History, Session, Device, etc.) to support multiple worker types, introduce a clean WorkerType/ApplicationType/TaskType model, and design simple job APIs so different worker pools can pull work and report results. In practice, that means adding a generic job/worker abstraction on top of what’s already there, not rewriting from scratch: the Spring app remains the single orchestrator, and emulator and headless services become pluggable executors behind it.
From there, you’ll implement and integrate the first headless worker service – likely using Playwright or a similar framework. This worker will fetch jobs from the control plane, run automated interaction and content workflows (e.g. posting, engagement, scripted journeys) for specific accounts, and push results back into the same execution history the emulator jobs use. You’ll also help wire in the initial CRM/funnel logic so multi-step outreach sequences can be expressed as data and emitted as jobs to the right worker type, respecting per-account limits and anti-detection constraints.
In parallel, you’ll help design the foundations of an AI-driven video pipeline that turns long-form source material into high-volume short-form assets. You won’t be training models from scratch, but you will be responsible for integrating with external AI services (for script generation, voiceover text, captioning, clip selection, etc.), defining the data contracts, and wiring the output back into the control plane as structured jobs. The goal is to treat “generate clips + copy + metadata for distribution” as just another worker pipeline the orchestrator can schedule, monitor, and feed into the posting and engagement workflows.
We’re looking for someone who is very strong in Java and Spring Boot, comfortable taking over and refactoring an existing backend, and experienced with distributed job/worker systems (job tables, status, retries, idempotency, simple REST or queue-based dispatch). You should be fluent with relational databases, REST API design, and running services on Linux (Ubuntu), and at least conversant with containerization (Docker, Kubernetes or similar). Hands-on familiarity with UI automation tools (Appium, Selenium, Playwright, Puppeteer, etc.) is important: you don’t have to be a career QA engineer, but you should understand the realities of flakiness, selectors, waits, and running many parallel sessions.
It’s a big plus if you’ve worked on high-scale marketing or growth automation, anti-detection/device farms, or CRM/funnel engines that coordinate multi-touch sequences over time. This role suits someone who enjoys taking a messy but valuable codebase, truly understanding it, and turning it into a coherent platform with clear contracts and extensible architecture. You’ll be expected to think in terms of control planes, worker pools, and future expansion (new platforms, new worker types) while still delivering quickly on immediate milestones.
If owning the brain of a real-world, high-volume automation platform sounds exciting, and you’re confident in your ability to drive a Java/Spring control plane from “partially built” to “running real campaigns,” we’d like to talk. Send a brief overview of your most relevant backend/platform work (especially job/worker systems or automation platforms), links to repos or case studies if you have them, and a short note on how you’d approach the first 4–6 weeks in this role.
PDF to Word Yearbook Name Extraction Category: Data Entry, Data Extraction, Editing, Microsoft Office, Microsoft Word, PDF, Proofreading, Word Budget: $750 - $1500 USD
Thai Entrance Face Image Dataset Category: Data Collection, Data Management, Event Planning, Face Recognition, Image Processing, Lighting Design, Photography, Video Production Budget: $250 - $750 USD
29-Dec-2025 23:03 GMT
CNC CAD Drawing for Foam Category: 3D CAD, 3D Drafting, AutoCAD, CAD / CAM, CNC, CNC Programming, Product Design, Solidworks Budget: $30 - $250 AUD
Curate 1000-Image Research Dataset Category: BeautifulSoup, Data Collection, Data Management, Data Mining, Python, Research, Selenium, Web Scraping Budget: ₹12500 - ₹37500 INR
29-Dec-2025 22:53 GMT
Natural Disaster Crisis Communications PlanS Category: Brand Management, Business Writing, Content Writing, Management, Media Relations, Project Management, Public Relations, Research, Research Writing, Social Media Management Budget: $250 - $750 CAD
29-Dec-2025 22:53 GMT
Short-Form Instagram/TikTok Videos -- 2 Category: Adobe Premiere Pro, After Effects, Final Cut Pro, Motion Graphics, Social Media Marketing, Video Editing, Video Production, Video Services Budget: $250 - $750 USD
