Founding Machine Learning Engineer

About UTony UTony is building a behavioral security system that helps financial organizations reduce the risk of Account Takeover attacks and suspicious login attempts. Our system analyzes user behavior to determine whether a given session matches the user's typical usage pattern or shows signs of suspicious activity. The problem and core solution have already been validated, and we are now focused on building the ML foundation of the product and taking it to the next stage. About The Role We are looking for a Founding ML Engineer who will take full technical ownership of our Machine Learning direction, from behavioral data and feature engineering to model development, evaluation, and production deployment. This is not a task-based role. We are looking for an independent, hands-on engineer who can take a semi-structured problem, explore the data, formulate hypotheses, design and run experiments, evaluate results, and take solutions into production. In this role, you will have a direct impact on how we build our ML systems, which behavioral signals we use, how we evaluate detection quality, and how we integrate these systems into the product. What You'll Work On Build and develop behavioral ML systemsWork on classification and anomaly detection problemsAnalyze behavioral and temporal/sequential dataBuild features from raw behavioral signalsDevelop, evaluate, and improve ML modelsDesign and run ML experimentsMeasure detection quality across different user and attack scenariosWork with precision, recall, false positives, false negatives, and decision thresholdsIdentify and prevent data leakageInvestigate model robustness and failure casesTake ML models into productionMonitor model performance and continuously improve modelsBuild and optimize low-latency, real-time inference systemsIndependently shape the key directions of our ML architecture and technical solutions What We're Looking For We are looking for an engineer who has already worked on real-world ML problems and understands the entire journey from data and experimentation to production. Strong knowledge of ML algorithms is important, but so is the ability to make independent decisions, understand trade-offs, and choose the approach that provides the most value for the actual product problem. Core requirements: 3+ years of hands-on Machine Learning experience preferredStrong candidates with 2+ years of relevant production experience are also encouraged to applyStrong Python skillsStrong understanding of Machine Learning fundamentalsExperience with classification problemsExperience with anomaly detection or similar problemsStrong experience with pandas, NumPy, and scikit-learnExperience with XGBoost or LightGBMStrong understanding of model evaluation, validation, precision, recall, thresholds, and data leakageExperience taking ML solutions beyond notebooks and prototypes into productionGood understanding of PostgreSQL / SQLExperience with Git and DockerStrong problem-solving skills and a high level of autonomy Nice To Have Experience with: Behavioral analyticsFraud detectionCybersecurityBot detectionAccount Takeover detectionUser/session modelingTime-series or sequence modelingOnline learningModel explainabilityReal-time ML systemsML latency optimizationExperience with Kubernetes, AWS / GCP / Azure, PyTorch, Kafka, Redis, or JavaScript / Node.js is also welcome but not required. What We Offer This is a full-time Founding ML Engineer position with real ownership of the Machine Learning direction. You will work directly with the CEO and Head of Engineering and have the opportunity to shape the core technical direction of our ML systems from an early stage. The position also includes a profit-sharing opportunity, with the terms to be discussed and defined in the employment agreement. Hiring Process Our hiring process typically includes: CV reviewTechnical interviewPractical discussion or task, if necessaryFinal discussion We are more interested in what you have actually built, how you approach ambiguous problems, and how you measure the results of your solutions than in the number of technologies listed on your CV. If you can take an ambiguous ML problem, explore the data, formulate hypotheses, design and evaluate experiments, build solutions, and take them into production, this role may be a strong fit for you.