Dynamic document intelligence, policy management, and predictive risk detection for Docufy.ai - ON-1208

Genre de projet: Recherche
Discipline(s) souhaitée(s): Génie - informatique / électrique, Génie, Informatique, Sciences mathématiques
Entreprise: Karbon Digital Ltd.
Durée du projet: Flexible
Date souhaitée de début: Dès que possible
Langue exigée: Anglais
Emplacement(s): Toronto, ON, Canada
Nombre de postes: 2
Niveau de scolarité désiré: MaîtriseDoctoratRecherche postdoctoraleNouvelle diplômée/nouveau diplômé
Ouvert aux candidatures de personnes inscrites à un établissement à l’extérieur du Canada: No

Au sujet de l’entreprise: 

Karbon Digital Ltd. is a Canadian technology company focused on artificial intelligence, intelligent document processing, regulatory technology, policy management, governance automation, and enterprise risk solutions. Its Docufy.ai platform applies multimodal large language models, semantic classification pipelines, and adaptive workflow automation to convert complex documents into structured, actionable, and auditable outputs.

The company’s mission is to advance trusted, privacy-preserving, and commercially deployable AI systems for regulated and documentation-intensive industries. Karbon Digital Ltd. supports use cases across compliance operations, policy interpretation, governance reporting, contract and document review, risk triage, and enterprise workflow automation.

The company is developing new Docufy.ai capabilities for dynamic document intelligence, policy-change management, predictive risk detection, multilingual content interpretation, and multi-model AI orchestration. Its research and development focus includes evidence traceability, explainability, confidence scoring, audit logs, human review workflows, and invention-ready technical assets that can support protectable intellectual property and Canadian AI commercialization.

Veuillez décrire le projet.: 

The project will extend Docufy.ai, Karbon Digital Ltd.’s intelligent document processing platform, by developing a Dynamic Document Intelligence, Policy Management, and Predictive Risk Detection module for multi-industry, multimodal, multi-model, and multilingual operating environments.

The project will focus on continuous policy and document-change monitoring, real-time interpretation of regulatory or organizational policy updates, multilingual content analysis, downstream impact assessment, and coordinated workflow implementation. The intended outcome is a validated research-to-product prototype that can classify complex documents, identify policy-relevant changes, detect latent compliance conflicts, recognize early risk indicators, and produce auditable outputs with lineage, confidence scoring, verification evidence, and human-review controls.

The intern will conduct applied research, review technical literature, design semantic marker methods, evaluate multimodal and multi-model AI approaches, develop document classification and risk-detection pipelines, test multilingual policy interpretation techniques, and support benchmarking across industry scenarios. The intern will also perform error analysis, ablation testing, reproducibility documentation, and technical reporting. Research outputs may support invention-disclosure-ready innovations involving semantic marker analysis, agentic orchestration, policy-lineage modeling, and compliance-grade verification protocols.

The project is a product, process, and organizational innovation initiative. It combines artificial intelligence, natural language processing, information governance, regulatory technology, enterprise automation, and responsible AI. Research methods may include retrieval-augmented generation, semantic search, embeddings, document structure analysis, multimodal content processing, language-model evaluation, confidence calibration, policy-risk taxonomy design, human-in-the-loop validation, and auditable workflow modeling.

Expertise ou compétences exigées: 

The ideal researcher should have expertise in computer science, software engineering, artificial intelligence, machine learning, natural language processing, data science, computational linguistics, information systems, or a related engineering or mathematical science discipline.

Experience with large language models, multimodal AI, multilingual NLP, retrieval-augmented generation, semantic search, embeddings, document classification, information extraction, model evaluation, and applied AI prototyping is preferred. Strong Python programming skills are required. Experience with APIs, data pipelines, cloud platforms, vector databases, evaluation frameworks, and software engineering practices would be valuable.

Familiarity with responsible AI, explainability, privacy-preserving AI, governance, compliance technology, auditability, policy management, regulatory workflows, and enterprise risk management is an asset. Knowledge of legal technology, financial services, healthcare, insurance, public sector, life sciences, or other regulated industries would strengthen the candidate’s fit.

The researcher should be comfortable designing experiments, evaluating model performance, documenting reproducible findings, performing error analysis, and translating research outputs into product-ready technical artifacts. Experience with academic writing, technical documentation, and invention-oriented research reporting is preferred.