AI Engineer

Gigi is the agentic operating system of enterprise media buying. Our first product is the AI media manager for Amazon DSP. Since launching it in summer 2025, we've grown to manage hundreds of millions in advertising spend on behalf of the largest agencies in the world.

We're backed by top venture investors Golden Ventures and Aperiam Ventures. Our founding team has spent 15+ years building in Amazon Ads and ecommerce, including a prior $100M+ exit.

We're hiring an AI Agent Engineer to join our engineering team in Toronto and own end-to-end development of the agents and workflows behind Gigi.

This is fundamentally a software engineering role. Much of the work will feel familiar: designing APIs, modelling data, integrating services, writing tests, debugging production issues, and making architectural decisions. Your focus will be how LLMs reason, use tools, retain context, and take actions on behalf of customers—and the software that makes those actions reliable.

You do not need to be an AI researcher or have years of professional LLM experience. Strong software engineers who have started building with LLMs and agents, and want to take that work further, should apply. You'll use tools such as Codex, Claude Code, and Cursor as a core part of how you work, while building across agent logic, backend services, data, and the customer-facing product.

As an early employee, you will have a meaningful role in shaping our product, engineering culture, and company.

WHAT YOU'LL DO

• Evals. Build a sophisticated eval framework across every Gigi use case, quantitative and qualitative, tied to real outcomes (spend, ROAS, margin, profit). Deterministic checks, LLM-as-judge graders calibrated against human review, trajectory-level evals for multi-step runs, and replay against historical account data. No model, prompt, or harness change ships without passing them.
• Context engineering. What goes into the context window on every call: advertiser context, account state, AMC results, retail signals, agency best practices, and prior decisions. You'll own retrieval, compaction for long-running trajectories, and cache-aware prompt design.
• Memory. Short-term, long-term, and cross-session memory for every brand, agency, and media buyer. The hard part is deciding what is worth remembering, so you'll build the agents that distill approvals, rejections, and outcomes into durable memory and handle conflicts and staleness. Memory stays model-agnostic and strictly separated per customer, including customers that require no training on their data.
• Agent behavior. Sub-agents, how tools are presented to the agent, and the line between what the LLM reasons about and what deterministic code executes. Risk and confidence scoring, guardrails, and how Gigi earns more autonomy with each customer. When a new model makes a workaround unnecessary, you delete it.
• Continual learning and the moat. The loop around the agent: capture traces, extract learnings, update context, memory, harness, or model weights, measure, repeat. Every action should leave Gigi smarter and more personalized for that customer, and outcomes across all customers should make every recommendation better.
• Model routing. Pick the right model for every call on price, performance, and speed, inside the harness, with fallbacks across providers. Stay model-agnostic, evaluate new models within hours of release, and own model spend.
• Post-training. The strategy and the implementation. Start with narrow, high-volume tasks (classification, extraction, summarization, guardrails, routing) and move toward the core agent loop when the data supports it, using fine-tuning, preference optimization, distillation, and RL against outcome-based rewards. You'll build the training data from traces and make the build vs buy calls on training and serving with engineering leadership.

WHAT SUCCESS LOOKS LIKE

• First 30 days. You've shipped your first improvements to production, stood up trusted evals on our core workflows, and contributed meaningfully to our roadmap.
• Six months. Every change ships through evals. Context and memory measurably improve recommendation quality per customer. Routing has cut model cost per action without hurting quality.
• Twelve months. A post-trained open-weight model is running in production. The commerce P&L runs live for brands on outcome-based contracts.

WHAT YOU'LL BRING

• A track record of building and shipping production software used by real customers. We care more about what you have built and owned than an exact number of years.
• Strong software engineering experience, with depth in backend or full stack development. System design, APIs, data models, testing, debugging, and operating production software are central to this role.
• Hands-on experience building with LLMs or agents, whether at work, through open source, or in a substantial personal project. You can explain how it works, where it fails, and how you would improve it.
• Experience using, or a strong desire to become highly effective with, AI coding tools such as Codex, Claude Code, Cursor, or similar tools. You do not need to be an expert on day one, but you should be excited to use them every day—and strong enough technically to review their work, reject weak abstractions, catch missing edge cases, and remain accountable for what ships.
• Curiosity about how LLMs change software products and the way software gets built. You should be excited to develop good judgment around where AI belongs—and where conventional software is the better answer.
• Judgment over specification. You can make a reasoned call, ship it, and iterate fast when it's wrong.
• Speed and ownership. LLMs changed what “fast” means for a small team. You should be comfortable using AI coding agents to move quickly while maintaining a high bar for architecture, correctness, and maintainability.

WHAT YOU DON'T NEED

• You do not need an AI research background or experience training foundation models. We build products around existing models, combining them with tools, data, business logic, interfaces, and production infrastructure.
• You do not need prior experience with every agent framework or concept we use. We expect strong engineers to learn LangGraph, agent orchestration, evaluations, retrieval, and tool use on the job.
• You do not need prior advertising or Amazon DSP experience. Our team will help you learn the domain.

BONUS POINTS IF

• You've built with LLMs in production: evals, guardrails, tool use, retrieval, agent orchestration, cost and latency work.
• You've built evaluation datasets or used production traces and customer feedback to improve an agent's behavior.
• You've worked in advertising or ecommerce.
• You've built real-time data pipelines at meaningful volume, or been early enough at a startup to know what the first 20 people actually do.
• You've operated production systems where correctness, permissions, auditability, or customer trust mattered.

THE HARD PARTS

We'd rather you know this before you apply.

• We work hard. We are trying to build a generational company, and early employees should expect intensity, ambiguity, and high standards.
• Cash compensation is competitive for our stage, but we will not win every late-stage salary comparison. The upside is meaningful ownership and equity in a company growing quickly.
• We throw out working code when the underlying abstraction is wrong. Some of what you build will not survive, and you'll be the one who calls it.
• Nobody is siloed. You'll write frontend, debug backend, think through product, join customer calls, and solve whatever problem is blocking the customer.

COMPENSATION AND EQUITY

• Base salary: $150k to $210k depending on experience level.
• Meaningful equity. We want to be as generous with equity as possible, and top performers can expect equity top-ups early and often.

LIFE AT GIGI

• In-office culture. We work together downtown Toronto (min 4x days per week), but makers (i.e. engineers) can choose to work where they are most productive on Wednesdays. Note: we're moving into a new office right across from Union.
• Catered lunch every day from Toronto's best spots. Current rotation: Impact Kitchen, iQ Food Co., ChopHop, and Cumbrae's.
• Real snacks: Greek yogurt, Mateína yerba mate, Barebells protein bars, Midday Squares. Tell us what you want and we'll add it.
• Top-of-the-line Apple equipment.
• Comprehensive health and dental coverage.