Draft:PromptQL

Submission declined on 10 December 2025 by Qcne (talk).

This draft's references do not show that the subject qualifies for a Wikipedia article. In summary, the draft needs multiple published sources that are:

in-depth (not just passing mentions about the subject)
reliable
secondary
independent of the subject

Make sure you add references that meet these criteria before resubmitting. Learn about mistakes to avoid when addressing this issue. If no additional references exist, the subject is not suitable for Wikipedia.

If you would like to continue working on the submission, click on the "Edit" tab at the top of the window.
If you have not resolved the issues listed above, your draft will be declined again and potentially deleted.
If you need extra help, please ask us a question at the AfC Help Desk or get live help from experienced editors.
Please do not remove reviewer comments or this notice until the submission is accepted.

Where to get help

If you need help editing or submitting your draft, please ask us a question at the AfC Help Desk or get live help from experienced editors. These venues are only for help with editing and the submission process, not to get reviews.
If you need feedback on your draft, or if the review is taking a lot of time, you can try asking for help on the talk page of a relevant WikiProject. Some WikiProjects are more active than others so a speedy reply is not guaranteed.

How to improve a draft

Wikipedia:Contributing to Wikipedia – a basic overview on how to edit Wikipedia.
Help:Wikitext – how to use the markup
Help:Referencing for beginners – how to include references
Wikipedia:Article development – how to develop your article
Wikipedia:Writing better articles – how to improve your article
Wikipedia:Verifiability – make sure your article includes reliable third-party sources

You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed article.

Improving your odds of a speedy review

To improve your odds of a faster review, tag your draft with relevant WikiProject tags using the button below. This will let reviewers know a new draft has been submitted in their area of interest. For instance, if you wrote about a female astronomer, you would want to add the Biography, Astronomy, and Women scientists tags.

Add tags to your draft

Editor resources

Easy tools: Citation bot (help) | Advanced: Fix bare URLs

Declined by Qcne 13 days ago. Last edited by Qcne 13 days ago. Reviewer: Inform author.

Resubmit

Please note that if the issues are not fixed, the draft will be declined again.

PromptQL

PromptQL is an artificial intelligence software platform created by the team behind Hasura, known for the open-source Hasura GraphQL Engine.^[1]^[2] The platform provides a natural-language interface for integrating large language models with enterprise data, focusing on accuracy and determinism for analytics and automation applications.^[3]

Overview

PromptQL functions as a data access layer for enterprise environments, positioned as a successor to GraphQL for interacting with business data through natural language.^[4] The platform uses an architectural approach that separates query planning from execution. Large language models generate structured query plans in a domain-specific language, while a deterministic runtime executes these plans outside the LLM context.^[5]

The system incorporates what the company calls an "agentic semantic layer" that learns organizational context, business rules, and terminology over time.^[6] This semantic layer encodes business definitions and logic as reusable, versioned metadata enforced at query time through GraphQL, SQL, or API calls.^[7] The platform connects to data warehouses, transactional databases, SaaS applications, and APIs using a federated query engine derived from Hasura's Data Delivery Network.^[8]

History

Early Development and Launch

Hasura publicly introduced PromptQL in June 2025, describing it as the company's first new data-access product since launching its GraphQL Engine in 2018.^[9]^[10]

The platform emerged from Hasura's experience with the GraphQL Engine, which had been downloaded over 500 million times globally.^[11] Hasura achieved unicorn status following a $100 million Series C funding round in February 2022, valuing the company at $1 billion.^[12]^[13] Hasura was founded in 2017 by Tanmai Gopal and Rajoshi Ghosh.^[14]

Academic Partnership and Benchmark Development

In June 2025, PromptQL announced a research collaboration with the University of California, Berkeley's EPIC Data Lab to develop a comprehensive benchmark for evaluating AI data agents in enterprise environments.^[15] The partnership, led by Professor Aditya Parameswaran, aims to address what Parameswaran described as the "1% problem" - the tendency of existing benchmarks to focus on capabilities relevant to technology giants while overlooking the complexity of real-world enterprise data.^[16]

The collaboration incorporates datasets from PromptQL's deployments across telecommunications, healthcare, finance, retail, and anti-money laundering sectors.^[17] The benchmark beta was scheduled for release in late 2025.^[18]

Consulting Services Launch

In September 2025, PromptQL launched an AI consulting practice offering direct access to its engineering team at $900 per hour.^[19] This positioned the company in competition with traditional management consulting firms by deploying engineers who built the platform.^[20]^[21] The consulting model targets Fortune 500 companies.^[22]

Technology

Architecture and Execution Model

PromptQL's core architecture decouples the planning and execution phases of data operations.^[23] When a user poses a natural language query, a foundational LLM generates a multi-step query plan expressed in PromptQL's domain-specific language.^[24] This DSL supports three categories of operations: data retrieval, data computation and aggregation, and semantic operations.^[25] The query plan is then executed deterministically in a runtime environment that operates outside the LLM's context window.^[26]

The platform stores intermediate results in structured artifacts that can be referenced in subsequent queries, enabling complex, multi-step workflows without consuming LLM context tokens.^[27]

Agentic Semantic Layer

The agentic semantic layer serves as a continuously evolving knowledge base that captures organizational semantics, procedural rules, and contextual relationships.^[28] Unlike traditional semantic layers requiring manual metadata curation, PromptQL's semantic layer learns from user interactions and corrections.^[29] When the system encounters ambiguous business terms or makes errors, user feedback is incorporated into the metadata.^[30]

The semantic layer is implemented as version-controlled YAML metadata that defines business concepts, metrics with their formulas and aggregation rules, relationships between entities, and access control policies.^[31]^[32] This metadata is bootstrapped from existing sources, including database schemas, internal wikis, and business intelligence definitions, then refined through usage.^[33]

Integration and Deployment

PromptQL integrates with Hasura's Data Delivery Network to provide federated access across heterogeneous data sources.^[34] The platform supports multiple interfaces, including a programmable Agent API for building AI assistants, a Program API for embedding workflows, and implementations of the Model Context Protocol for integration with AI development tools.^[35]^[36]

The system offers multiple deployment options, including a managed cloud service, virtual private cloud deployment, and self-hosted installations.^[37] All deployment models support enterprise security requirements, including role-based access control, row-level and column-level security policies, and audit logging.^[38]^[39]

Performance Claims

PromptQL reports achieving over 90% accuracy on complex datasets through its learning-based approach that captures business ontology and procedural semantics.^[40] The company has published benchmark comparisons demonstrating what it characterizes as improvements over traditional tool-calling and retrieval-augmented generation approaches.^[41]^[42]

Industry Position

PromptQL operates in the emerging category of enterprise AI platforms focusing on structured data access.^[43] The platform competes with text-to-SQL solutions from vendors such as Vanna.ai, SQLAI.ai, and offerings from cloud providers like Google Cloud's BigQuery SQL generation and Microsoft's SQL Server 2025 AI features.^[44]^[45]^[46]

Reception

Coverage of PromptQL has primarily focused on its enterprise-oriented approach and consulting model. VentureBeat profiled the company's consulting practice in September 2025, highlighting its positioning against traditional consulting firms.^[47] The UC Berkeley partnership garnered attention from technology industry observers.^[48]

Relationship to Hasura

PromptQL is developed by the creators of Hasura and is featured in Hasura's product lineup alongside the company's GraphQL and API offerings.^[49]^[50] The platform builds on Hasura's Data Delivery Network infrastructure and shares its federated query engine architecture.^[51] Tanmai Gopal serves as CEO of both Hasura and PromptQL.^[52]^[53]

Hasura raised a total of $136.5 million in funding through its Series C round in February 2022, led by Greenoaks with participation from Nexus Venture Partners, Lightspeed Venture Partners, Vertex Ventures, and STRIVE.^[54]^[55] The company operates with offices in San Francisco and Bengaluru.^[56]

References

External links

[Official website](https://promptql.io/)
[PromptQL Documentation](https://github.com/hasura/promptql-docs)
[UC Berkeley EPIC Data Lab](https://epic.berkeley.edu/)

Category:Artificial intelligence Category:Software companies of the United States Category:Companies based in San Francisco Category:Database management systems Category:Natural language processing Category:2025 establishments in the United States Category:Software using the Apache license

[1] ttps://hasura.io/resources/promptql-100-percent-accurate-ai-agent-on-your-data

[2] ttps://hasura.io/blog/from-graphql-to-promptql-a-new-chapter-begins

[3] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[4] ttps://hasura.io/blog/from-graphql-to-promptql-a-new-chapter-begins

[5] ttps://promptql.io/blog/how-promptql-achieves-100-accuracy-for-ai-on-enterprise-data

[6] ttps://promptql.io/blog/the-ultimate-guide-to-semantic-layers-for-ai

[7] ttps://hasura.io/blog/rethinking-the-semantic-layer-for-the-ai-era

[8] ttps://promptql.io/blog/bridging-ai-and-enterprise-data-with-promptql-x-mcp

[9] ttps://hasura.io/blog/from-graphql-to-promptql-a-new-chapter-begins

[10] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[11] ttps://hasura.io/blog/from-graphql-to-promptql-a-new-chapter-begins

[12] ttps://techcrunch.com/2022/02/22/graphql-developer-platform-hasura-raises-100m-series-c/

[13] ttps://www.businesswire.com/news/home/20220222005420/en/Hasura-Announces-$100M-in-Series-C-Funding-at-a-$1B-Valuation-to-Make-GraphQL-Available-to-Everyone

[14] ttps://hasura.io/blog/hasura-year-one-50def1cc7b73

[15] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[16] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[17] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[18] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[19] ttps://venturebeat.com/ai/promptqls-usd900-hour-ai-engineers-are-coming-for-mckinseys-ai-business

[20] ttps://venturebeat.com/ai/promptqls-usd900-hour-ai-engineers-are-coming-for-mckinseys-ai-business

[21] ttps://promptql.io/blog/engineering-led-consulting-will-shape-future-of-enterprise-ai

[22] ttps://venturebeat.com/ai/promptqls-usd900-hour-ai-engineers-are-coming-for-mckinseys-ai-business

[23] ttps://promptql.io/blog/how-promptql-achieves-100-accuracy-for-ai-on-enterprise-data

[24] ttps://promptql.io/blog/how-promptql-achieves-100-accuracy-for-ai-on-enterprise-data

[25] ttps://www.youtube.com/watch?v=1nOTQsfe1RU

[26] ttps://promptql.io/blog/how-promptql-achieves-100-accuracy-for-ai-on-enterprise-data

[27] ttps://promptql.io/blog/building-a-powerful-customer-support-ai-assistant-with-promptql-in-5-minutes

[28] ttps://promptql.io/blog/the-ultimate-guide-to-semantic-layers-for-ai

[29] ttps://promptql.io

[30] ttps://promptql.io

[31] ttps://promptql.io/blog/the-ultimate-guide-to-semantic-layers-for-ai

[32] ttps://hasura.io/blog/rethinking-the-semantic-layer-for-the-ai-era

[33] ttps://promptql.io/blog/the-ultimate-guide-to-semantic-layers-for-ai

[34] ttps://promptql.io/blog/bridging-ai-and-enterprise-data-with-promptql-x-mcp

[35] ttps://promptql.io/blog/bridging-ai-and-enterprise-data-with-promptql-x-mcp

[36] ttps://github.com/hasura/promptql-mcp

[37] ttps://platform.softwareone.com/product/promptql/PCP-3328-9237

[38] ttps://promptql.io/blog/bridging-ai-and-enterprise-data-with-promptql-x-mcp

[39] ttps://platform.softwareone.com/product/promptql/PCP-3328-9237

[40] ttps://www.globenewswire.com/news-release/2025/06/04/3093929/0/en/PromptQL-Partners-with-UC-Berkeley-to-Develop-New-Data-Agent-Benchmark-for-Reliability-of-Enterprise-AI-Agents.html

[41] ttps://hasura.io/resources/promptql-100-percent-accurate-ai-agent-on-your-data

[42] ttps://promptql.io/blog/how-promptql-achieves-100-accuracy-for-ai-on-enterprise-data

[43] ttps://platform.softwareone.com/product/promptql/PCP-3328-9237

[44] ttps://www.bytebase.com/blog/top-text-to-sql-query-tools/

[45] ttps://www.getgalaxy.io/resources/best-text-to-sql-tools-2025

[46] ttps://www.microsoft.com/en-us/sql-server/blog/2025/05/19/announcing-sql-server-2025-preview-the-ai-ready-enterprise-database-from-ground-to-cloud/

[47] ttps://venturebeat.com/ai/promptqls-usd900-hour-ai-engineers-are-coming-for-mckinseys-ai-business

[48] ttps://techedgeai.com/news/promptql-uc-berkeley-partner-on-enterprise-ai-data-agent-benchmark/

[49] ttps://hasura.io/blog/from-graphql-to-promptql-a-new-chapter-begins

[50] ttps://hasura.io

[51] ttps://promptql.io/blog/bridging-ai-and-enterprise-data-with-promptql-x-mcp

[52] ttps://venturebeat.com/ai/promptqls-usd900-hour-ai-engineers-are-coming-for-mckinseys-ai-business

[53] ttps://councils.forbes.com/profile/Tanmai-Gopal-Co-founder-CEO-PromptQL/c366490c-54a1-45eb-bbd0-4fd937e05c1f

[54] ttps://techcrunch.com/2022/02/22/graphql-developer-platform-hasura-raises-100m-series-c/

[55] ttps://www.businesswire.com/news/home/20220222005420/en/Hasura-Announces-$100M-in-Series-C-Funding-at-a-$1B-Valuation-to-Make-GraphQL-Available-to-Everyone

[56] ttps://golden.com/wiki/Hasura-YX39VD9

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]