Example Analysis
Machine Learning Engineer — Recommendations Platform
A real, unedited output from the Job Fit Evaluator (generated 2026-07-05). Nothing here was touched up — including the grade.
Strong Python and PyTorch skills but lacks large‑scale recommendation engineering experience
Satori brings solid Python and PyTorch production experience and has built model evaluation and latency‑optimization pipelines, which align partially with the ML engineering requirements. However, he lacks the 4+ years of large‑scale recommendation system engineering, distributed training, and real‑time feature store experience that the role demands.
Strengths
At HeartStamp he rebuilt the Flux LoRA training pipeline and created LangSmith prompt regression suites and golden datasets for model evaluation, using PyTorch in production.
Gaps & Weaknesses
Honest assessment — these are real gaps, not spin.
Relevant Case Studies
From Broken Pipeline to Board Seat
Diagnosed LoRA training failures, pivoted the company to a model-agnostic prompting system using art historical vocabulary, guided MVP launch, and joined as Board Member.
From Interrogation to Creative Session
Redesigned Stampy's interaction model from an average 8-question interrogation (96% of users bounced before seeing a card) into a 4-step creative session that 87% of users complete.
When the Tool Is Its Own Proof
Designed and shipped 'specify' — an open-source Claude skill that walks any agent task through the Five Pillars of Specification Engineering — and published it as a reusable workflow anyone can deploy.
One Model Is Not Enough
Published Phase2S: a full AI coding agent harness with 29 built-in skills, cross-model adversarial review, a 'dark factory' for autonomous spec execution, and 1,191 tests. Works on your existing ChatGPT subscription — no API key required.
The job description it evaluated
Machine Learning Engineer — Recommendations Platform We're looking for an ML Engineer to build and productionize recommendation models serving 20M daily users. What you'll do: - Train, evaluate, and deploy ranking and retrieval models (PyTorch) - Build feature pipelines and serving infrastructure (Python, Spark, Airflow) - Own model monitoring, drift detection, and A/B experiment analysis - Optimize inference latency and GPU serving costs at scale - Work with data engineering on real-time feature stores Requirements: - 4+ years production ML engineering experience - Strong Python; hands-on PyTorch or TensorFlow in production - Experience with distributed training and model serving infrastructure - Solid grasp of experiment design and offline/online evaluation
Paste your own job description and watch it work in real time.
Run your own analysis