Observed Signal · Jul 22, 2026 · Technical Release · Source: DEV Community · Impact: 2/5 · Sentiment: Positive
Onibex Builds Governed Text-to-SQL Agent for SAP
Onibex published ASK (Agentic Semantic Knowledge), an open platform that converts natural language into governed SQL over SAP data by using a curated semantic layer. ASK represents business concepts as YAML “Data Products” organized in a Bronze/Silver/Gold medallion model: Bronze exposes raw SAP tables, Silver defines join logic and field roles, and Gold provides denormalized analytics for low-latency queries. The platform offers three query engines (Flash, Smart, Precise) that trade off latency, cost, and determinism. Disambiguation is handled via a three-level system indexed in OpenSearch and managed through an ASK Configuration App (React SPA). ASK can produce governed artifacts (downloadable Excel reports) and the semantic layer specification, platform code, and manual are published on GitHub.
The release demonstrates a governed, deterministic approach to LLM-to-SQL over complex enterprise schemas (SAP), which is relevant to enterprise analytics and data governance but is niche to SAP/enterprise data use cases rather than broadly industry-shifting.
Track SAP Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Onibex published ASK (Agentic Semantic Knowledge), an open platform that maps natural language to governed SQL over SAP data.
- ASK uses a YAML-based semantic layer called Data Products and applies a Bronze/Silver/Gold medallion model for schema abstraction.
- The platform provides three query engines — Flash (1 LLM call, ~15s), Smart (2 LLM calls, ~40s), and Precise (3 LLM calls, ~60s) — to trade off speed, cost, and determinism.
- Disambiguation uses a three-level system backed by a semantic dictionary index in OpenSearch and is manageable via an ASK Configuration App (React SPA).
- The semantic layer specification, platform code, and full manual are published on GitHub.
Connected Companies & Entities
2 Entities mapped“But SAP HANA schemas are not simple....”
“The semantic layer specification, the platform code, and the full manual are published on GitHub....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Actian Launches AI Analyst for Smarter Conversational Analytics
Actian, the data and AI division of HCLSoftware, introduced Actian AI Analyst (formerly Wobby), a conversational analytics solution that uses a Semantic Knowledge Graph and a new Steward Agent to build and maintain a governed semantic layer. The product combines AI-driven natural-language querying with deterministic semantics and a constrained execution engine to reduce hallucinations, provide full traceability of reasoning steps, and maintain consistent business logic. Actian AI Analyst integrates with collaboration tools such as Slack and Microsoft Teams, preserves threaded conversational context, and can be paired with the Actian Data Intelligence Platform to create a closed-loop governed analytics system. The company presents the solution as aimed at scaling analytics access for business users while ensuring auditable, trusted insights.
AI Agents Need a Semantic Layer
The article argues that giving AI agents direct text-to-SQL access to data warehouses exposes a lack of shared business semantics, causing inconsistent answers, missing access controls, and no auditability. A semantic layer (an intermediary metadata/metrics layer) solves these problems by exposing governed metric definitions, enforcing row-level security and multi-tenancy, providing an API (MCP/tool APIs) rather than raw DB connections, and enabling metrics-as-code workflows (YAML + Git). The piece distinguishes 'agentic semantic layers'—designed for programmatic agent consumption—from traditional BI semantic layers, lists operational requirements (MCP support, pre-aggregation, schema-as-code, warehouse coverage), and names existing vendors/tools. It promotes discovery and query APIs and recommends versioned, governed metric definitions to ensure consistent, auditable results across dashboards, agents, and APIs.
Knowi Launches Enterprise Data Agents with Built-in Private AI
Knowi announced a new set of enterprise AI agents for its analytics platform that automate end-to-end BI workflows from a single prompt. The release includes 20+ purpose-built agents, an orchestrator and an MCP (Model Context Protocol) server that lets MCP-compatible AI clients (e.g., Claude) call Knowi’s querying, dashboarding and alerting tools. Knowi runs its own AI infrastructure so queries and schemas are processed without routing data to third-party LLMs; the product supports on-premises deployment, enforces enterprise access controls and row-level permissions, and is SOC 2 Type II certified with HIPAA-compatible configurations. Agents are embeddable via SSO/API and integrate with Slack and Microsoft Teams. The platform connects to 70+ data sources (SQL, NoSQL, REST, SaaS, documents) and claims cross-join queries without ETL or a data warehouse.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.