BENCH+ PROFILE

Eugene

ML Engineer
NLP, LLM evaluation, content classification, and production ML pipelines

Production-oriented ML engineer with long-running experience in NLP, classification, OCR, anti-fraud scoring, model monitoring, and data pipeline implementation. Best aligned with applied ML delivery where Python, PyTorch, orchestration, evaluation, and production feedback loops are central.

BEST-FIT USE CASE

Applied ML engineering for NLP or classification systems, including dataset preparation, model training, evaluation harnesses, deployment support, monitoring, and iterative improvement in production-oriented environments.

Not the primary fit: Roles centered primarily on cloud data-platform configuration, CRM specialization, or standalone data engineering should include a focused hands-on validation of those responsibilities before selection.

EVIDENCE HIGHLIGHTS

What supports the assessment

Designed, built, and operated a production LLM quality evaluation framework used for release decisions across classification and search-quality workflows.

Implemented automated regression evaluation, statistical gating, structured-output scoring, and production-traffic drift checks for model and prompt changes.

Designed and implemented automated annotation pipelines using LLM-based workflows integrated with distributed data infrastructure and workflow orchestration.

Built ML pipelines for unwanted-content filtering, including labeling processes, model training and distillation, production integration, monitoring, and A/B testing.

Delivered applied ML solutions across recommendation, insurance estimation, logistics optimization, and enterprise data-processing contexts.

Built real-time anti-fraud scoring and feature-engineering pipelines for financial transaction streams, with monitoring and rollout processes.

Developed OCR and text-recognition pipelines for structured and unstructured documents in healthcare, construction, and document-management contexts.

WORKING STYLE

How they show up

Works with Scrum or Agile-style task flow.
Clarifies tasks with colleagues, leads, and task authors when requirements need refinement.
Has participated in code review processes.
Has experience working in small teams and in production-oriented ML workflows.
Has direct client communication experience according to interview notes.

KEY STRENGTHS

Implemented production-oriented ML workflows for content classification, search-quality filtering, anti-fraud scoring, OCR, and recommendation-related use cases.
Built evaluation and regression workflows for LLM quality assessment, including structured scoring, repeated sampling, statistical checks, and drift-triggered re-evaluation.
Hands-on with end-to-end ML lifecycle activities: dataset construction, model training, deployment integration, monitoring, A/B testing, and retraining policies.
Experience with streaming and behavioral feature engineering for financial transaction risk scoring.
Worked across multiple applied domains including finance, healthcare, logistics, e-commerce, media, document processing, and enterprise software.
Has operated in small engineering teams and has experience clarifying tasks with leads, colleagues, and task authors.
Has led small-to-medium engineering groups in ML delivery contexts according to the provided career history.

PRIMARY SKILLS

PythonPyTorchMachine learningNLPLLM evaluationModel monitoringAirflowSpark

ADDITIONAL TECHNOLOGIES

SQLscikit-learnCatBoostXGBoostpandasDockerKubernetesMLflowOCRA/B testingETL pipelinesCloud data pipelines

INDUSTRY EXPERIENCE

AI / DataEnterprise SoftwareHealthcareLogisticsMediaFintechE-commerce

ROLE MATCH

ML EngineerStrong

Data / ML EngineerPartial