Sr. Full Stack Engineer

KeyBank•For Those Who Work At Home, GA

5d•$80,000 - $150,000•Hybrid

About The Position

KeyCorp is seeking a specialist to design, train, and operationalize intelligent document-processing solutions. The role combines applied AI and machine learning with cloud data engineering, secure information handling, and configurable business-rule automation. Its focus is transforming unstructured and semi-structured documents into accurate, auditable, actionable data products. The position involves document classification and extraction, model training and evaluation, document preprocessing and validation, human-in-the-loop review, and configuration-driven data validation. It also encompasses secure cloud processing, data pipelines, API and system integration, operational monitoring, and documentation. The engineer will work with business subject-matter experts to define fields, thresholds, exceptions, and acceptance criteria, and communicate technical concepts to both technical and non-technical stakeholders. The role is based in Georgia for employees who work at home. Key prioritizes in-office presence while providing flexible options when a role can be performed effectively in a mobile environment. This position is not eligible for employment visa sponsorship for non-U.S. citizens. Compensation is a base salary of $80,000 to $150,000 annually, with eligibility for incentive compensation that may include production, commission, and/or discretionary incentives. The posting references a benefits list but does not enumerate the benefits offered.

Requirements

Bachelor's degree in Information Systems, Computer Science, Data Analytics, or a related technical field.
Hands-on experience delivering document-processing or information-extraction solutions in a regulated industry.
Experience handling sensitive or confidential information and enterprise security controls.
Familiarity with cloud-based machine learning and data platforms.
Knowledge of financial-services, legal, healthcare, insurance, or other document-intensive workflows.
Experience designing exception-based workflows that reduce manual review while preserving quality and control.
Experience with intelligent document processing (IDP), optical character recognition (OCR), and document classification.
Experience training, evaluating, and tuning document extraction models, including for handwritten content.
Python development experience for data processing, automation, model integration, and reusable components.
Experience with cloud object storage, SQL-based analytical platforms, production data pipelines, and batch or event-driven processing.
Knowledge of configuration-driven validation using YAML or similar formats.
Knowledge of data normalization, reconciliation, tolerance handling, and comparison against authoritative sources.
Experience with encryption, decryption, sensitive-data controls, auditability, and secure cloud processing.
Knowledge of API and system integration for document ingestion, result publishing, and downstream workflows.
Understanding of supervised model training, labeling strategies, evaluation datasets, confidence thresholds, and error analysis.
Practical experience with document AI platforms or comparable machine learning services.
Experience applying large language models to document understanding, unstructured content analysis, summarization, or reasoning.
Knowledge of responsible AI practices, including explainability, traceability, quality controls, and review escalation.
Experience with workflow orchestration, retries, monitoring, operational support, CI/CD, source control, automated testing, environment promotion, and infrastructure as code.
Ability to translate review procedures into technical requirements and configurable validation rules, collaborate with subject-matter experts, and communicate clearly with technical and non-technical stakeholders.

Responsibilities

Design, train, and operationalize intelligent document-processing solutions.
Develop, train, evaluate, and tune document extraction models for structured, semi-structured, and handwritten content.
Implement entity and field extraction, confidence scoring, quality metrics, and model accuracy assessment.
Build document preprocessing, page-level validation, exception detection, and human-in-the-loop review patterns.
Develop Python data-processing and automation solutions, model integrations, and reusable components.
Build cloud-native data pipelines using object storage, analytical databases, batch processing, and job orchestration.
Implement configuration-driven validation rules using YAML or similar formats to separate business rules from application code.
Normalize and reconcile data, handle tolerances, and compare extracted information with authoritative data sources.
Apply encryption, decryption, sensitive-data controls, auditability, and secure cloud processing.
Integrate APIs and systems to ingest documents, publish results, and support downstream workflows.
Apply machine learning and AI approaches to document understanding, choosing appropriately among OCR, rules-based logic, machine learning, and generative AI.
Implement workflow orchestration, retries, monitoring, and operational support for batch and event-driven processing.
Design data models for extracted values, confidence scores, validation results, and processing history.
Use CI/CD, source control, automated testing, environment promotion, and infrastructure-as-code practices.
Troubleshoot and provide observability across model, application, data, and integration layers.
Translate document-review procedures into technical requirements and configurable validation rules.
Collaborate with business subject-matter experts to define fields, thresholds, exceptions, and acceptance criteria.
Document model behavior, processing logic, decision outcomes, and operational procedures.
Communicate AI and data-processing concepts to technical and non-technical stakeholders.
Develop scalable, reusable solutions rather than one-off document automations.