Explore enterprise products and AI capabilities built for industry.
Let’s talk about your businessProducts, integrations and partnershipsBring AI into the enterprise.
Automated data-warehouse builder
Automated high-quality datasets
AI appliance for business decisions
Opens in a new tabDedicated offline AI bidding appliance
Connect AI to products and operations.
Explore the interfaces and integration paths for our AI capabilities.
Models, foundational frameworks and the engineering behind AI systems.
Explore collaboration across products, technology and delivery.
Frameworks, tools and examples for developers to understand and reuse.
Updates, technical thinking and industry observations from IDENIFE.
Our focus on education, talent and sharing knowledge.
ID Optima
IDENIFE’s first fully automated high-quality dataset factory
Turn existing material and governed assets into datasets for AI, as well as business metrics and data marts. Powered by ID Axis, Optima connects requirements, processing, validation and version delivery in one production flow.
Documents · Records · Governed assets
Parse · Clean · Map · Validate
Data, definitions and quality evidence, together.
AI Ready · Data processing for models
Prepare data for RAG retrieval and model fine-tuning: parse, deduplicate, mask and structure content, retain provenance and access rules, then validate it against agreed criteria. Deliver the content, structure and evidence that the intended AI workflow requires.
Identify content and source structure
Deduplicate and mask sensitive content
Create chunks or training samples
Check structure, citations and access
Datasets, metadata and acceptance reports
Two uses. Two processing paths.
Retrieval needs complete semantic passages, together with provenance and access rules. Optima preserves the content and its conditions of use in the same dataset.
Product material · PDF
Record the equipment model, fault message and time of occurrence, and submit the details to the service team.
Disconnect power and confirm the equipment has stopped. Record inspection items and abnormalities after maintenance.
Parse headings, sections and body text to organize useful content.
Keep a complete topic in each passage to retain context.
Link source locations and access rules; check duplicates and sensitive content.
Record the equipment model, fault message and time of occurrence, and submit the details to the service team.
Disconnect power and confirm the equipment has stopped. Record inspection items and abnormalities after maintenance.
Metric generation · Data processing for business
Identify metrics and dimensions from requirements or report templates, interpret their context and map them to existing assets. Generate metric definitions, transformation SQL and data marts for downstream analytics and reporting systems.
Select a metric to inspect its definition.
| Metric | This month | Last month |
|---|---|---|
| — | — | |
| — | — |
The template defines what data is needed. Figures are not report results.
SUM(order_amount)The template calls for monthly sales. Map the sales amount and order date to produce a monthly aggregation from the governed order asset.
Use metric names, row and column context, and time conditions to form requirements that people can review.
Connect source fields, aggregations and dimensions, translating the metric requirement into processing logic.
People confirm refund timing or consolidation rules. Production branches that do not depend on those decisions can continue.
For data governance delivery teams
Extend existing governance work with AI-ready corpora, training samples and metric datasets. Give teams more room to focus on business definitions and delivery quality, with less repetitive mapping, transformation development and acceptance documentation.
Build retrieval corpora and training data on existing assets to support knowledge bases, fine-tuning and AI applications.
Connect requirements, field mapping and SQL generation in one flow, reducing manual translation and handoffs between stages.
Isolate decisions that need confirmation while independent branches continue. Collect production records and acceptance evidence with each version.
| Aspect | Common metric workflows | Optima dataset production |
|---|---|---|
| Starting point | Delivery teams translate business requirements into metric models, calculation rules and field relationships, then configure tasks. | Interpret goals from requirements or templates and propose metrics and mappings for business confirmation. |
| Execution | Execute configured tasks through SQL, visual pipelines and scheduling. | ID Axis decomposes goals and coordinates tools, connecting mapping, generated processing and validation. |
| Delivery scope | Focus on metric models, computed results and query services, supported by quality controls. | Produce AI corpora, training samples and metric datasets, with metadata, processing logic and acceptance evidence. |
Compared with common configuration-driven delivery workflows. Efficiency comes from fewer repeated configurations, handoffs and waits; actual gains depend on source quality, business complexity and acceptance criteria.
Technology foundation
Purpose, intended users, output formats and acceptance criteria form a target contract. The confirmed agreement guides production and the checks on the final output.
ID Axis.
ID Axis breaks data production goals into coordinated responsibilities, connecting model reasoning, tool calls and execution feedback. Optima applies this framework from requirements to deliverable datasets.
HuanYu provides semantic understanding and analytical reasoning to identify the structure, context and acceptance requirements of retrieval corpora, training samples and metric datasets.
ID Axis orchestrates parsing, mapping, transformation and validation tasks, invokes the appropriate tools and uses execution feedback to progress production.
Ambiguous definitions and mappings are raised for confirmation while independent branches continue. Decisions, production records and version outputs stay in the same project.
Quality and version delivery
High quality is checked against explicit rules. A version stays pending while blocking rules remain unmet and proceeds to release and delivery once the requirements are satisfied.
Check citations and access for retrieval, structure and masking for training, and fields, calculations and definitions for metrics. Each rule has an explicit result.
A candidate version brings data, metadata and acceptance reports together. Offline packages preserve a snapshot from creation; later updates produce a new package.
Illustration · knowledge retrieval dataset
Passages stay linked to their source.
Check each passage for its content, source document and section reference, with the required field types.
The delivery audience needs confirmation.
The source is restricted to the service team. A broader delivery audience requires permission to be confirmed first.
Keep a record of how data was handled.
Apply the agreed masking rules to personal details, retaining the affected fields and processing results for review.
Unresolved issues stay visible.Blocking checks prevent release until the corresponding issue is resolved and validated again.
ID Optima
Start with assets built in Omnia or existing material. Optima produces AI-ready datasets and business metrics that take your data into new applications.