From raw datato regulatory-gradereporting.

No-code data pipelines

Compose Source, Select, Join and Aggregate objects visually. Each step is a versioned, reusable object that writes to its own partitioned table — no glue scripts, no hidden state.

Durable workflow orchestration

Pipelines run as Temporal workflows with automatic retries, real-time SSE status, and per-variant isolation. Long-running, fault-tolerant, and observable from the first node to the last.

Real-time streaming

Stream from Kafka or watched directories through Select, Join and Aggregate on Apache Spark Structured Streaming — watermarked, stateful, and idempotent, with live throughput metrics.

REST API in & out

Ingest from any REST API with API-key, OAuth2, Basic or Bearer auth, smart pagination and rate-limiting — then publish governed results as secure, keyed REST endpoints.

AI pipeline generation

Describe a transformation in plain language; DGenXT drafts the pipeline — sources, joins, aggregates, reports — to review and run.

Multi-database

Dialect adapters target Postgres, Oracle, MySQL, MariaDB and MSSQL — the same pipeline, portable across estates.

Data quality, enforced

DataQuality and Dcheck objects validate records against rules before they reach a report — bad data caught at the source.

Six report formats

Excel, CSV, XML, TXT, FreeHand and XBRL — generated from the same governed data, with templating, multi-sheet styling and scheduling.

Maker-checker governance

Four-eyes approval on record adjustments and multi-level sign-off on reports, with enforced segregation of duties and an append-only audit trail.

Ingest from any API

• Four auth schemes — API key, OAuth2 client-credentials, HTTP Basic and Bearer — credentials stored RSA-encrypted, never in plaintext.
• Smart pagination — Offset, page, cursor and token strategies, with watermark-based incremental pulls so you only fetch what changed.
• Rate-limited & resilient — Per-API requests-per-minute with token-bucket limiting and exponential backoff on 429/503.
• SSRF-protected — Outbound targets validated against private, loopback and link-local ranges before any call is made.

Publish as an API

• Keyed public endpoints — Publish any pipeline output as a REST endpoint secured by SHA-256 consumer keys with optional expiry.
• Column-level security — An allow-list controls exactly which columns are exposed — no accidental data leakage.
• Partitioned & filtered — Consumers query by business date and column filters; results paginate with dialect-aware SQL.
• Full access logging — Every request is logged with IP, key hint, rows returned, duration and status for audit.

Excel, done properly

• Apache POI engine — Multi-sheet workbooks with fonts, fills, borders, merged cells, embedded images and freeze-pane headers.
• Configurable styling — Per-report header and data styles, number and date formats, and auto-sized columns.
• Templated & scheduled — Drive layouts from templates and emit on a workflow schedule, from the same governed data.

Validation you can see

• Rule-based checks — DataQuality and Dcheck objects enforce null, range, regex and cross-field rules before data is reported.
• Five XBRL validators — Schema, calculation, dimensional, formula and regulator-specific checks gate every regulatory filing.
• Highlighted in the viewer — The report workspace surfaces validation results as colored pass / warn / fail badges and per-fact validation tables — so exceptions are obvious before you submit.