AI/ML Engineer with Azure AI Document

Role: AI/ML Engineer with Azure AI Document

Location: Edina, MN/ Chicago, IL/ Irving, TX

Long Term

Type: Contract

Client: CITIUS

Job Description:

Senior Azure AI Document Intelligence Engineer design, train, evaluate, and continuously improve the document-processing models used for supply-chain contracts and invoices.
This engineer will own extraction quality and be directly accountable for achieving the project’s 98–99% accuracy objectives.

Key Responsibilities

· Analyze invoice and contract variations across suppliers, countries, languages, formats, and scan-quality levels.

· Define the document taxonomy and extraction schema for:

· Invoice headers

· Contract metadata

· Purchase-order references

· Supplier and customer information

· Dates, currencies, taxes, discounts, and totals

· Payment terms, renewal dates, obligations, and termination clauses

· Tables and invoice line items

· Evaluate Azure AI Document Intelligence prebuilt invoice, layout, custom neural, custom template, classification, and composed-model capabilities.

· Build and train custom document classifiers and extraction models.

· Prepare, label, clean, balance, and version training and evaluation datasets.

· Implement document preprocessing for rotation, skew, noise, resolution, page separation, and scan-quality issues.

· Develop confidence-scoring and validation strategies using field-level and OCR-level confidence signals.

· Implement deterministic validation rules, including:

o Subtotal, tax, and total reconciliation

o Currency and date validation

o Purchase-order and supplier matching

o Duplicate-document detection

o Required-field and cross-field consistency checks

· Establish confidence thresholds and human-review rules for uncertain extractions.

· Perform error analysis by document type, supplier, field, language, and scan quality.

· Create automated evaluation pipelines reporting exact-match accuracy, precision, recall, F1 score, false-positive rates, and false-negative rates.

· Monitor model drift and retrain models when document formats or business requirements change.

· Document model versions, training data, experiments, limitations, and release decisions.

· Collaborate with supply-chain, procurement, accounts-payable, legal, and engineering teams.

Required Qualifications

· Bachelor’s or master’s degree in Computer Science, Data Science, Artificial Intelligence, Engineering, or a related discipline.

· 4+ years of experience in machine learning, OCR, computer vision, NLP, or intelligent document processing.

· Hands-on experience with Azure AI Document Intelligence or a comparable enterprise document-processing platform.

· Strong Python development skills.

· Experience with REST APIs, JSON, Azure SDKs, and asynchronous processing.

· Experience training and evaluating document classification and field-extraction models.

· Strong understanding of accuracy, precision, recall, F1 score, confidence calibration, and test-set design.

· Experience extracting complex tables and variable-length line items.

· Experience creating labeled datasets and maintaining data-quality standards.

· Familiarity with invoices, purchase orders, contracts, or supply-chain documents.

· Strong analytical, debugging, and technical-documentation skills.

Preferred Qualifications

· Experience with Azure Machine Learning, MLflow, Azure AI Search, or Azure OpenAI.

· Experience with multilingual documents.

· Familiarity with contract clause extraction and legal-document processing.

· Knowledge of ERP or procurement systems such as SAP, Oracle, Dynamics 365, Coupa, or Ariba.

· Experience implementing active learning, model-drift monitoring, and human-in-the-loop workflows.

· Knowledge of data privacy and document-retention requirements.

· Success Measures

o ≥99% exact-match accuracy for approved critical fields.

o ≥98% exact-match accuracy across all agreed fields.

· Accuracy maintained across suppliers, formats, and document-quality categories.

· Measurable reduction in manual corrections and review time.

· Documented improvement process for fields that miss their thresholds.

· No production model deployed without passing the approved evaluation suite.