Original job description
Developer (Production Engineering and AI Platform)
About Elumai
Elumai is a precision learning and AI tutoring platform. It builds a model of how each learner thinks and adapts every interaction to match. We are built in Canada and work with workforce and trades programs, enterprise learning teams and post-secondary institutions. We are a small team preparing for launch with our first pilot partners, and we work without layers of management.
Purpose
Help take Elumai from pilot to production. Keep the platform fast, stable and reliable for learners, many of them in locations with limited internet.
Grow into a long-term technical partner across future contracts and builds: new courses, new clients, and scaling our capacity for serving learners.
Responsibilities
First priorities
Uptime and resilience.
Prevent and recover from outages: cloud services, GPU servers, deploys
Monitoring, alerting, health checks after deploys, safe rollbacks.
Speed
Find and fix the causes of slow responses: page loads, API latency, AI response time.
Make the platform work well on low-bandwidth, high-latency connections.
Web app quality
Stability, performance and polish of the learner interface on desktop and mobile, including accessibility.
Release discipline
Automated tests, staged rollouts, and checking real behaviour before and after each release.
Longer-term responsibilities
Course creation pipeline. Learn our curriculum ingestion pipeline, which turns source documents (curricula, microcredentials, exam materials) into structured course content.
Content validation. Check that structured content is retrieved correctly, served by the AI tutor, and displayed properly in the frontend.
Structured data. Design and validate structured JSON flowing between AI models, backend services and the UI, much of it generated on demand.
New builds. Support new courses, clients and features; take on parts of the technical lead's workload.
Experience
Must-have skills
Google Cloud in production: Cloud Run, Cloud Build / CI-CD, Cloud SQL (PostgreSQL), Compute Engine, logging and monitoring, Secret Manager.
Backend: Python (FastAPI or similar), async services, REST/streaming APIs.
Frontend: TypeScript/React (Next.js), with an eye for UI/UX, responsive design and performance.
Reliability and performance: profiling and diagnosing latency end to end (browser to CDN to server to database to model).
Structured data: JSON schemas, validation, data pipelines.
Testing: pytest, browser automation (e.g., Playwright), regression testing.
Strongly preferred
AI in production: serving LLMs (e.g., vLLM on GPUs), RAG/retrieval (pgvector, embeddings), and evaluating AI output quality.
CDN/edge experience (Cloudflare)
Working effectively with AI coding assistants as part of the daily workflow.
Experience in instructional design, teaching or education
Nice to haves
Ed-tech or learning-platform experience; LMS integrations.
Familiarity with privacy and data-sovereignty principles
Experience building for low-connectivity regions.
How we work
A small team with a lot of ownership: comfortable diagnosing problems across the whole stack, not just one layer.
Evidence-driven: measure, reproduce, fix, verify.
Employment details
Employment type: 6 month contract until March 31st, 2027
Hours: 20-30 hrs/week
Compensation: $70–$100 per hour
Location: Remote, Must be able to work Eastern Standard Time Zone hours. Candidates must be based in Canada.
Start date: Mid-November
More about this job
Responsibilities
Maintain and improve the platform’s uptime, resilience, speed, and learner-facing web experience across desktop and mobile, including for users with limited connectivity. Build reliable release and testing practices, support course-content pipelines and AI content validation, and contribute to new courses, clients, and features.
Requirements
Candidates need production experience with Google Cloud services, Python backend development, and TypeScript/React frontend development, along with skills in end-to-end performance diagnosis, structured data validation, and automated testing. Strongly preferred experience includes production LLM serving, retrieval-augmented generation, CDN/edge services, AI coding assistants, and education or instructional design.
Skills
- Google Cloud
- Cloud Run
- Cloud Build
- Cloud SQL
- PostgreSQL
- Compute Engine
- Python
- FastAPI
- Async Services
- REST APIs
- TypeScript
- React
- Next.js
- LLM Serving
- RAG and Retrieval
- JSON Schema Validation
Remote locations (standardized)
Visa sponsorship
Not detected in the job text
Categories
- Technology
- Software
- Education
- Engineering
Keywords
- Google Cloud
- Cloud Run
- Cloud Build
- CI/CD
- Cloud SQL
- PostgreSQL
- Compute Engine
- Secret Manager
- Python
- FastAPI
- REST APIs
- Streaming APIs
- TypeScript
- React
- Next.js
- LLM Serving
- vLLM
- GPU Servers
- RAG
- pgvector
- Embeddings
- JSON Schemas
- Pytest
- Playwright
- Cloudflare
- Monitoring and Alerting
- Performance Profiling
- Accessibility
- Low-Bandwidth Optimization
- Instructional Design
