Why this page exists
The purpose of this page is to separate longer-term ideas from the capabilities in DataRead Free 0.1.0, DataRead Advanced 1.0.0 and DataRead Professional 2.0.0. The current products remain limited to their published release descriptions.
25 proposed Enterprise feature areas
The 25 areas below span data engineering, analytics, machine learning, language intelligence, model operations, governance and decision support. Related ideas are separated where they represent different stages or responsibilities. No delivery order or priority is committed.
Expanded regression, hypothesis testing and A/B testing
Proposed experimental and inferential analysis beyond the current Advanced edition's supported statistical workflows.
Advanced Analytics
Time-series forecasting and anomaly intelligence
Proposed analysis of sequential data, future-value estimates and unusual changes over time.
Forecasting
AutoML, hyperparameter tuning and model comparison
Proposed automated testing and ranking of candidate models and configurations; distinct from Professional's guided ML workflow.
Machine Learning
Neural networks and GPU-assisted training
Proposed neural-network workflows and optional acceleration on supported hardware.
Deep Learning
Advanced deep-learning architectures and computer vision
Proposed image and visual-data analysis using more advanced model architectures.
Deep Learning
Text analytics, language processing and sentiment analysis
Proposed extraction of themes, entities, classifications and sentiment from text data.
Language Intelligence
Connections to existing language models for data questions
Proposed connections that let users ask natural-language questions about supported datasets and analysis results.
Generative AI
Local language models, document Q&A and RAG
Proposed local-model options and retrieval-based question answering over approved documents and project materials.
Generative AI
Language-model training and fine-tuning
Proposed tools to prepare training material and adapt supported language models for specific tasks.
Generative AI
Reusable analysis workflows and experiment tracking
Proposed saved, repeatable analytical workflows with recorded configurations, runs and experiment outcomes.
Automation
Model registry, versioning, APIs and deployment tools
Proposed central model records, version control, controlled interfaces and deployment support.
Model Operations
SQL, database and expanded data-source connectors
Proposed connections to additional structured data stores and enterprise source systems.
Data Integration
Model explainability, governance and audit controls
Proposed model explanations, documented approvals, traceable changes and governance records.
Trust and Governance
AI agents and automated multi-step analysis
Proposed task-oriented agents that coordinate multiple approved analysis and data-preparation steps.
Advanced AI
Synthetic data generation and recommendation systems
Proposed privacy-aware artificial datasets and systems that identify relevant options or recommendations.
Advanced AI
DataRead Vault: project catalogue, snapshots and backups
Proposed local-first project organisation, recoverable snapshots and managed backup records.
Data Management
Automated data-pipeline orchestration
Proposed scheduling of imports and transformations, dependency management, failure handling and automatic retries.
Data Engineering
Data observability and end-to-end lineage
Proposed monitoring of freshness, schema changes, quality signals and the path data takes through projects and outputs.
Data Reliability
Enterprise data catalogue and semantic layer
Proposed searchable data definitions, business glossary, ownership metadata and consistent shared metric meanings.
Data Management
Enterprise privacy and security management
Proposed role-based access, sensitive-data classification, masking, encryption and configurable retention controls.
Security and Privacy
Production ML monitoring and MLOps
Proposed monitoring of deployed models for drift, performance changes, latency and operational issues after release.
Model Operations
Machine-learning feature store
Proposed reusable, documented and versioned model input features for consistent training and prediction workflows.
Machine Learning
Real-time data streaming and intelligence
Proposed processing of continuously arriving data with event-triggered analysis, alerts or predictions.
Data Engineering
Decision intelligence, simulation and optimisation
Proposed what-if scenarios and constrained optimisation to compare alternatives and support operational decisions.
Decision Support
Knowledge graphs and graph-enhanced AI
Proposed modelling of entities and their relationships to enrich search, discovery and context-aware analysis.
Knowledge Intelligence
How these ideas differ from the current editions
- Advanced 1.0.0 already provides its documented statistical profiles, distributions, category/group/relationship analysis, graphics and diagnostics. Enterprise proposals extend beyond those released analytical tools into expanded experimental analysis, forecasting and decision optimisation.
- Professional 2.0.0 already provides guided baseline classification, regression and clustering workflows, Batch Processing and supported model evaluation. Proposed AutoML automates candidate selection and tuning; deep learning, feature stores and production MLOps are separate proposed capabilities.
- Data engineering proposals would add scheduled orchestration, streaming, observability, lineage, catalogues and connectors. These are not part of the current single-file or batch-preparation product boundaries.
- Model lifecycle proposals separate training and current model evaluation from post-deployment monitoring, drift detection, reusable feature management, versioning and controlled operations.
- Language and knowledge proposals include text analysis, language-model connections, local document Q&A, RAG and knowledge graphs. None is represented as a current Free, Advanced or Professional feature.
- Governance and security proposals would explore access controls, sensitive-data handling, privacy management, auditability and documented governance. They are not current enterprise security certifications or controls.
- DataRead Vault remains a proposed local-first project catalogue, snapshot and backup concept. This page does not promise hosted storage, cloud collaboration or a managed cloud service.
Non-binding roadmap notice
DataRead Enterprise is a proposed future edition and is not currently released, available to purchase, order, download, test or receive support for. The 25 areas below are exploratory concepts only. Their names, scope, architecture, feasibility and priorities may change. Any item may be revised, combined, postponed, replaced, reduced, cancelled or never developed. No release date, price, availability, support commitment, upgrade path, licence inclusion or delivery guarantee is offered. These descriptions are working concepts, not a product specification.
Current released editions
Only DataRead Free 0.1.0, DataRead Advanced 1.0.0 and DataRead Professional 2.0.0 are presented as the current released editions on this website. Check their individual pages and exact release packages for current scope.