AI Engineer — LLM Systems & Agents

# AI Engineer — LLM Systems & Agents *KM Capital Corp* · Concord, Ontario · Full-time · On-site in Concord to start, remote flexibility after · Works directly with the Chief AI Officer --- ## Job details •⁠ ⁠*Job type:* Full-time, permanent •⁠ ⁠*Pay:* $55,000–$65,000 per year, plus benefits •⁠ ⁠*Expected hours:* 40 per week •⁠ ⁠*Schedule:* Monday to Friday, day shift •⁠ ⁠*Work location:* On-site in Concord, Ontario to start. Once you know the business and the systems, we're open to remote. •⁠ ⁠*Expected start date:* Immediate --- ## Read this part first You already know if this is you. You've been building with language models since the day you got your hands on them. You have projects nobody asked you to make. You lose evenings to something you can't stop thinking about, and you've probably been told you need a few more years before anyone will let you near anything real. We're not going to tell you that. We want one person with more drive than résumé — someone who can't stop building and has nowhere good to point it. We're not looking to train you. We're looking for someone to build alongside. ## Who we are KM Capital is a Canadian specialty finance company. We fund small and mid-sized businesses through merchant cash advances and manage that portfolio ourselves. We are further into automation than anyone in our industry — and it's live, not planned. We build our own software instead of buying it. What's live right now: •⁠ ⁠Real-time conversational voice agents over telephony — outbound calls handled end to end, with live transcription, tool calls mid-conversation and structured outcomes written straight back to our systems •⁠ ⁠Document ingestion that classifies, extracts and routes what comes in, without a human reading it first •⁠ ⁠Retrieval across our own corpus, with a governed context layer that decides what a model is allowed to see before it answers •⁠ ⁠Multi-step agent workflows with tool calling, error handling and monitoring, running unattended •⁠ ⁠Event-driven integrations across credit, banking and identity data services •⁠ ⁠A major platform build in its final stretch before rollout That is what's in production today, and it makes us the leader in our industry. There's a long list of what comes next — that's what you'd be here for. ## The bar Four things. That's the whole list. •⁠ ⁠*You can code.* Fluent in at least one modern language — TypeScript/JavaScript or Python is ideal, but if you're strong elsewhere and can move, say so. •⁠ ⁠*You can read code you didn't write* and work out what it actually does. •⁠ ⁠*You build with models daily.* Claude Code, Cursor, the model APIs, agent frameworks, MCP — this is already how you work. •⁠ ⁠*You've shipped something real.* Something that exists, that runs, that someone could use. Size doesn't matter. Working does. Nobody arrives knowing everything. You should be the kind of person who closes the gap on their own, fast, without being walked through it. ## What you'll be building You're stepping straight into the final stages of a major CRM build — a large, live platform in its last stretch before full rollout. You'll be in it in week one, on real code, with real users waiting. After that: •⁠ ⁠Applications and internal tools powered by language models, across operations, finance, sales and customer communications •⁠ ⁠Agent workflows that take over manual process end to end •⁠ ⁠Voice and messaging systems that talk to real customers •⁠ ⁠Retrieval and document pipelines over our own data •⁠ ⁠Integrations connecting third-party data services into systems people use every day Real production systems, real users, real consequences — the thing side projects can never give you. The Chief AI Officer brings the vision and the problems worth solving. You build them, with real latitude in how. And you bring your own — if you think there's something we should build, make the case. Good ideas get built here regardless of who had them. ## Deliberately not required •⁠ ⁠*A degree.* We hire on what you've built and whether it works. Self-taught engineers are genuinely welcome here. •⁠ ⁠*Years of experience.* Show us the work instead. •⁠ ⁠*A perfect track record.* Half your projects being unfinished is normal for someone who builds constantly. What matters is carrying the ones that count all the way to done. ## Why you'd want this •⁠ ⁠*You'll ship.* What you build goes live and gets used, usually within days. •⁠ ⁠*You'll be heard.* Freedom in how things get built, and real influence on what gets built next. •⁠ ⁠*Direct access.* You work side by side with the Chief AI Officer. Decisions happen in a conversation, not a committee. •⁠ ⁠*You'll be properly resourced.* Tools, licences and compute are budgeted for. You won't be fighting for the things that make you effective. •⁠ ⁠*Flexibility, earned.* You'll be on-site at the start — there's no substitute for absorbing how this business actually works. After that, we're open to remote. •⁠ ⁠*You're first.* This team is being built right now, and whoever joins at the start has the most room to grow with it. ## Benefits •⁠ ⁠Extended health care •⁠ ⁠Dental care •⁠ ⁠Vision care •⁠ ⁠Paid time off •⁠ ⁠On-site parking •⁠ ⁠Catered lunch once a week •⁠ ⁠Company events, including summer cottage days •⁠ ⁠Casual dress ## How to apply Send us four things: 1.⁠ ⁠*Your resume.* 2.⁠ ⁠*A link to something you built* — running, or the repository, whichever shows it best. 3.⁠ ⁠*A few sentences on the hardest problem you hit building it, and how you solved it.* This is the part we read most closely. 4.⁠ ⁠*One thing you'd want to build here,* based on what you've read above. A cover letter is welcome if you have something to say — we'll read it. It just won't carry as much weight as the work. If your best work is under an employer agreement or otherwise can't be shared, say so. We'll sign a mutual NDA and find a way for you to show us. *Then we'll talk.* A short call, then a scheduled session where you walk us through something you built — your screen, your code, your decisions. It's a conversation, not a test. Bring the project you're proudest of. --- ## Summary *Job Type:* Full-time, permanent *Pay:* $55,000–$65,000 per year *Expected hours:* 40 per week *Schedule:* Monday to Friday, day shift *Education:* Not required *Experience:* Building software with language models — 1 year (preferred) *Language:* English (required) *Work authorization:* Must be legally entitled to work in Canada *Ability to commute:* Concord, ON — reliably commute or plan to relocate before starting (required for the initial on-site period) *Work Location:* In person to start, with remote flexibility once established ### Application questions 1.⁠ ⁠Share a link to something you have built — a live project, a demo, or a repository. 2.⁠ ⁠What was the hardest problem you hit building it, and how did you solve it? 3.⁠ ⁠What is one thing you would want to build here, based on this posting? 4.⁠ ⁠Which tools do you currently build with day to day? --- KM Capital Corp is an equal opportunity employer. Accommodations for applicants with disabilities are available on request at any stage of the recruitment process.