Open role
Lead AI/ML Engineer —
build the loop.
loomai sits at the intersection of professional services and product. We build continuously learning AI systems for enterprises in regulated industries. The edge is no longer in who has the most data — it's in who can prompt, fine-tune, evaluate, and orchestrate most effectively. We need someone who owns that end to end.
What We Build
A continuously learning decision loop — Read, Act, Observe, Tune.
The heart of our platform runs in production on real enterprise data across multiple engines simultaneously. This is what you'll own.
Ingest and structure signals from enterprise data sources in real time — the foundation every decision is built on.
Score, route, and execute decisions using a combination of ML classifiers, rule engines, and agentic frameworks.
Close cohorts, assemble ground truth from real outcomes, and build the feedback signal that makes the next cycle smarter.
Use LLMs to detect patterns, draft retraining proposals, and promote models safely — with a human in the gate where it matters.
You'll own the full stack: the data pipelines that feed it, the models that score it, the frameworks that deploy and promote them safely, and the LLM-powered layer that finds patterns and proposes what to change next. This runs on real enterprise data, serves real decisions, and gets measurably better over time.
How to Apply
Show us how you think.
We reviewed over 1,300 applications for our last hire. We're not doing that again. We ask for one thing up front: a short video. If it lands, we'll ask for more.
Respond to the prompt below in a video. No production value required — we're watching how you think.
Email your video, CV, and LinkedIn profile to careers@loomai.ai. Add one line on why you want to work with us.
If your video resonates, we may send a short async ML system design problem. 48 hours. We want to see how you approach a real, messy problem — not a whiteboard exercise.
A 45-minute conversation with Tricia and David. We'll go deep on your work, your thinking, and what you want to build.
The Video Prompt
One question. Two to three minutes. Your voice.
You've built a model that scores a decision — routes a customer, scores their fit, or composes an offer. It goes live. 60 days in, how do you know if it's working? How do you assemble ground truth, and how do you decide when and how to retrain?
No longer. Editing is a skill. If you need 10 minutes, that's a signal.
Don't script it. We want to hear how you actually think through ML system design.
Tell us what you'd actually do — including what you'd skip or simplify at an early stage.
Ground truth assembly, retraining triggers, deployment patterns. You've thought about these. Show us.
What We're Looking For
ML depth. Data instinct. Client presence.
Must-Haves
- 3–5 years of ML engineering or applied AI — production systems, not notebooks
- Python fluency; traditional ML (LightGBM, XGBoost, scikit-learn) and modern AI stacks (LangChain, LlamaIndex, OpenAI/Anthropic APIs)
- RAG, prompt engineering, evaluation harnesses, and/or fine-tuning on domain-specific data
- Agent orchestration — systems where models make, route, or compose decisions
- Data pipeline design at enterprise scale
- Full model lifecycle: training, evaluation, shadow, canary, A/B, ongoing tuning
- Client presence — hold your weight with domain experts, explain a model to a CFO
- Startup mentality — move fast, own your work, don't wait to be told
Strong Plus
- Salesforce ecosystem — Agentforce, Data Cloud, Einstein, Apex, Flow
- Financial services background
- Versioned policy/rule deployment patterns
- Product instinct — governable and scalable, not just functional
- Prior consulting or client-facing applied AI experience
What We Offer
Real ownership. Real clients. Real upside.
Ready? Send us your video.
Email your video link, CV, and LinkedIn profile to careers@loomai.ai. One line on why loom·ai.
Subject line: Lead AI/ML Engineer Application — [Your Name]