Full-Stack Developer | AWS, APIs & AI Coding Agents | $5/hr Starting Rate

Individual contractors only. Agencies need not apply. About Neo Doc Neo Doc is an AI-powered dental software platform supporting clinical workflows such as SOAP notes, periodontal charting and patient onboarding. We’re looking for a hands-on full-stack developer to help our existing engineering team implement features faster while maintaining reliable, secure code. This is an implementation-focused role. You must be comfortable writing and debugging code yourself, using advanced AI coding tools effectively, and independently checking their output. We value working features, reproducible tests and clear communication over activity metrics or large amounts of generated code. Compensation and growth The starting rate is $5 USD/hour, firm and non-negotiable. Please apply only if you are comfortable with this initial rate. We offer a performance-based path toward $10/hour over the first year: - Starting rate: $5.00/hour - After 30 days: eligible for $5.50/hour - After 90 days: eligible for $6.50/hour - After 180 days: eligible for $7.50/hour - After 270 days: eligible for $8.50/hour - After 365 days: eligible for $10.00/hour These are review opportunities, not guaranteed increases. Progress depends on consistently delivering accepted work, code quality, security, reliable testing, accurate time tracking, documentation and teamwork. Changes require written agreement and an updated Upwork contract. Hours and availability will be agreed before work starts. ***You must personally perform all work under this contract. No outsourcing, subcontracting, account sharing, or delegating work to other people is permitted. Approved AI coding and QA agents are encouraged, but you remain personally responsible for understanding, reviewing, testing, and delivering the work. What you’ll work on: - Build and improve frontend features and backend services in an existing application. - Develop secure AWS-backed functionality and troubleshoot infrastructure and application issues. - Build and test integrations between our app, middleware and third-party APIs, including dental practice-management systems such as Open Dental. - Use Postman to reproduce API issues, maintain reusable collections and validate success, error and authorization cases. - Turn Jira acceptance criteria into small, reviewable GitHub pull requests with test evidence. - Work with our lead developer, coordinate ownership and flag dependencies instead of duplicating or conflicting with existing work. Required technical skills - Strong full-stack development skills, including TypeScript/JavaScript, React/Next.js and backend/API development. - Practical AWS experience, including IAM/least privilege, networking/VPC concepts, secure secrets management, application hosting, databases, storage and logs/monitoring. - Strong understanding of REST APIs, authentication, authorization, webhooks, validation, retries, idempotency and error handling. - Proficiency with GitHub branches, pull requests, code review and CI/testing workflows. - Effective use of Jira and Confluence for concise, accurate technical documentation and progress tracking. AI-assisted development is a prerequisite You should already use Codex and/or Claude Code as part of your development workflow, not be learning the basics for the first time on this project. You must be able to: - Configure approved MCP connections for tools such as GitHub, Jira, Confluence and relevant development services, with appropriate access limits. - Maintain useful repository instructions, Markdown documentation, AGENTS.md/CLAUDE.md files and reusable coding skills. - Use coding, review and QA agents to implement, inspect and test changes efficiently. - Use current approved frontier models, such as Claude Opus 5.5 or GPT-6 Astra, or approved successors, through supported tooling. - Read, explain and debug the resulting code yourself. AI output must be reviewed, not blindly accepted. - Produce review-ready PRs with reproducible checks, without bypassing required human review or deployment approvals. Tool access, model usage and any associated costs must be agreed before work begins. Do not purchase services or add credentials or permissions without approval. How we define completed work: - The agreed acceptance criteria are implemented and demonstrated. - Relevant unit, API/integration and actual UI/workflow tests are run where applicable. - The PR identifies the exact commit tested, environment, results, remaining limitations and anything mocked or not tested. - Jira updates link to the PR and evidence. Confluence is updated when behavior, architecture or setup changes. - Setup instructions are reproducible by another developer. - A successful compile or an AI agent’s “passed” report alone is not acceptance. Use ticket-specific branches from the agreed baseline. No direct pushes to protected/shared release branches, unauthorized merges or production deployments. Report blockers promptly, with what failed and what is needed to proceed. Healthcare security and onboarding: We are building healthcare software that must meet HIPAA requirements. Healthcare or dental software experience is a strong advantage. - Use synthetic data for development, demonstrations and the hiring assessment. - Never put patient information, secrets or private credentials into unapproved AI tools, screenshots, logs or repositories. - Follow approved access, encryption, audit-logging and environment-separation practices. - Be willing to sign an NDA, confidentiality/IP-assignment agreements and other applicable onboarding/security documentation before receiving access. Time tracking and accountability You must be comfortable using Upwork’s time tracker and screenshot system on the approved computer where you actually perform the work. - Do not track time on one computer while working on an undisclosed second computer. - Approved cloud environments, VMs and remote development are allowed when accessed through the tracked workstation and clearly documented. - Use meaningful work-diary memos identifying the ticket and work performed. - No auto-clickers, simulated keystrokes or tools used to fabricate activity. - Keep unrelated personal activity and sensitive information out of captured work sessions. - Stop tracking during personal breaks or when not working. Any arrangements for unattended agent runs or other waiting time must be agreed in advance. - No account sharing or undisclosed subcontracting. Provide a short update each working day: what changed, PR/test links, blockers and next steps. Selection process Shortlisted candidates will complete a small paid practical assessment through Upwork, with scope and an hour cap agreed in advance. The assessment will use a limited sandbox or isolated portion of Neo Doc, synthetic data and temporary least-privilege access. No production access is required. You’ll implement a small feature or fix, submit a PR, demonstrate the result, provide tests and explain your technical decisions and AI-assisted workflow. We evaluate correctness, security, speed, maintainability and communication. All authorized assessment work is paid, whether or not you are selected for ongoing work. To apply Please answer: 1. Do you accept the firm $5/hour starting rate and performance-based review schedule? 2. Describe an AWS-backed application and API integration you personally built. What was your contribution? 3. Which AI coding tools, models and MCP integrations do you currently use? Give a concrete example of your workflow. 4. Share a permitted code/PR example showing implementation, tests and documentation. Do not share another client’s confidential material. 5. How do you verify AI-generated code and report tests that are mocked, blocked or not run? 6. What is your experience with healthcare software, Open Dental or other practice-management integrations? 7. Confirm your availability and willingness to complete a paid assessment, use accurate Upwork tracking and sign the required onboarding documents. We’re looking for someone who can move quickly, explain their work clearly and build software the team can confidently maintain.