AI Engineer — Agents & Backend

Kora Home AI; Toronto; Hybrid; CAD 110,000 - 150,000; Full time

Job summary

Kora Home AI is building an intelligent home assistant based in Toronto that helps busy families manage schedules and tasks. You will develop the core AI backend, focusing on prompts, tool integrations, memory, and reliable workflows to deliver proactive, context-aware assistance across multiple devices and surfaces.

The role combines back-end engineering with AI workflow design, enabling end-to-end orchestration from request to action, with a strong emphasis on dependable performance and user

Qualifications

• Strong Python skills and backend API development.
• Experience with LLM-powered apps beyond tutorials.
• Prompting, tool calling, structured outputs, and retrieval understanding.
• Experience shipping software and diagnosing production issues.
• Ability to work independently in a small team.

Responsibilities

• Own AI features from design to production deployment and iteration.
• Develop system prompts, context assembly, structured outputs, and tool‑calling behavior.
• Build Python backend services that connect models, data, and integrations.
• Design workflows that tolerate missing information and partial failures.
• Implement permissions and confirmation flows for family-automation scenarios.
• Build evaluations with realistic failure scenarios and metrics.
• Investigate agent traces and user feedback to improve workflows.

Skills

Python

Backend services

LLM prompts

Asynchronous tasks

Production debugging

Independent work

Small team

Tools

PostgreSQL

Firebase

GCP

Job description

Full-time · Toronto, Canada | Hybrid or remote

About Kora

Kora is a Toronto-based, VC-backed startup building an intelligent assistant for busy families. Our product combines a mobile app with a dedicated home display to help families manage schedules, chores, shopping, meals, and everyday responsibilities.

We are building an assistant that understands family context, remembers what matters, anticipates what needs attention, and takes action to help families get things done.

We are a small, hands-on team. You will work directly with the founders to build Kora's core intelligence and shape how the assistant behaves in everyday life.

The role

We are looking for an AI engineer with strong backend development skills to build and improve Kora's core assistant capabilities.

You will work on the full agent loop: understanding a request, gathering relevant context, deciding what to do, calling tools, checking results, and following through. Your work will include prompts, orchestration, memory, integrations, evaluations, and the services that run these workflows.

You should enjoy turning emerging AI capabilities into useful product features and making them dependable enough for real users.

What you’ll build

Agent reasoning and execution

The prompts, tool definitions, control flow, and execution harness that enable Kora to complete tasks across multiple steps.

Family context and memory

Systems that help Kora understand people, preferences, schedules, responsibilities, and ongoing commitments.

Proactive assistance

Workflows that detect changes, conflicts, preparation needs, and deadlines, then decide when and how to help.

Tools and integrations

Connections to calendars, tasks, shopping lists, email, and other services that let the assistant take action.

Conversational experiences

Support for requests through chat, voice, images, and forwarded information.

Reliable backend infrastructure

APIs, background jobs, event processing, and persistent workflow state.

Evaluation and observability

Tools to understand agent behavior, measure quality, investigate failures, and improve performance.

For example, when a family forwards a message about a changed volleyball practice, Kora should understand what changed, identify the relevant event, check for conflicts, and help coordinate the next steps with appropriate confirmation.

What you’ll do

• Own AI features from initial design and prototype through production deployment and iteration.
• Develop and refine system prompts, context assembly, structured outputs, and tool‑calling behavior.
• Build Python backend services that connect models, application data, and external integrations.
• Design workflows that can handle missing information, interruptions, retries, and partial failures.
• Implement permissions and confirmation flows so the assistant acts appropriately on a family’s behalf.
• Build evaluations using realistic scenarios and production failure cases.
• Investigate agent traces and user feedback to identify why a workflow succeeded or failed.
• Improve response quality, latency, and inference cost.
• Collaborate with mobile and device engineers to deliver a consistent assistant experience.
• Help decide when to use an LLM, a deterministic workflow, or a combination of both.

What we’re looking for

• Strong Python skills and experience building backend services, APIs, and database‑backed applications.
• Hands‑on experience building an LLM‑powered application beyond a basic tutorial.
• Practical understanding of prompting, tool calling, structured outputs, context management, and retrieval.
• Experience shipping software and diagnosing problems in production.
• Ability to reason about asynchronous tasks, application state, error handling, and data consistency.
• A systematic approach to testing AI behavior and evaluating whether changes actually improve the product.
• Product judgment: you care about whether the assistant solves the user's problem and earns their trust.
• Regular use of AI development tools, with ownership of the quality and correctness of the resulting code.
• Comfort working independently in a small team, clarifying ambiguous problems, and delivering complete features.

Helpful experience

• Production AI agents or workflows that execute actions through external tools.
• Agent memory, context engineering, retrieval systems, or knowledge graphs.
• Durable workflows, job queues, scheduling, and event‑driven systems.
• PostgreSQL, Firebase, GCP, or similar infrastructure.
• OAuth and integrations with calendar, email, or productivity services.
• Voice assistants, real‑time streaming, or multimodal applications.
• Agent evaluations, tracing, monitoring, and regression testing.
• Handling untrusted inputs, prompt injection, and permissions in AI applications.

We value strong engineering fundamentals, practical judgment, and evidence of building useful things. You do not need experience with every framework or technology listed.

• Build the core intelligence of a consumer AI product.
• Work directly with the founders and influence architecture and product behavior.
• Tackle real problems involving memory, reasoning, proactive assistance, and action.
• Connect AI capabilities to mobile software and a physical product in people's homes.
• Take substantial ownership and grow with an early‑stage, VC‑backed company.