Senior Software Engineer — AI-Assisted Development (Claude Code)
Must-have:AWSAISecurity
Scale an AI Trucking Platform from One Carrier to Many (AWS · AI · Claude Code) We are building an AI operations desk and TMS for trucking companies with 1 to 100 trucks in Canada and the USA. It reads a carrier’s mailboxes, GPS and documents, and turns every quote, load and problem into one tracked record.
Where we are: The product was built with Claude Code and runs live at our first carrier on AWS, with Claude, Gmail and Samsara integrated. Now we need to make it reliable, secure and ready for many carriers.
- Turning a single-carrier system into one that serves many carriers safely, with each company’s data fully separated and carrier rules handled as settings.
- Production on AWS: deploys, monitoring, alerts, backups, uptime and cost.
- AI quality: measuring accuracy, catching wrong answers, controlling cost per carrier, and testing changes before they ship.
- Integrations at scale: Gmail and Outlook mailboxes, Samsara and other GPS/ELD providers.
- Security: Google and Microsoft app verification, access control, audit logs, and carrier security reviews.
- Tests and code standards, so AI-written code can ship safely.
- Have 3+ years building production software, including time as an early engineer at a B2B software company that grew to many customers.
- Have run real systems on AWS and been on call for them.
- Have shipped LLM features to production and know how they fail.
- Use Claude Code or similar tools daily and can quickly spot what AI-written code got wrong.
- Write clearly and push back with reasons.
- Gmail API / Google OAuth in production, especially restricted-scope verification
- LLM structured outputs with real evaluations
- Logistics, accounting or other domains where a wrong number costs money