RelationalAI
RelationalAI builds a Snowflake-native decision intelligence platform using its proprietary Rel declarative language, semantic modeling, and neuro-symbolic reasoners to run AI agents inside customer Snowflake accounts, primarily serving telecom, financial services, and supply chain enterprises.
- Company typePrivate
- Founded2017
- HeadquartersBerkeley, United States
- Headcount101–250
- GTM typeB2B
- OfferingSoftware
What RelationalAI does
RelationalAI is a Berkeley-based software company founded in 2017 that builds a decision intelligence platform delivered as a Snowflake Native App and accessed through its proprietary declarative language, Rel, and its Python variant, PyRel. The platform combines a relational knowledge graph, a governed semantic-modeling application (Rel App), and four composable neuro-symbolic reasoners (rules, graph, predictive via graph neural networks, and prescriptive via constrained optimization), plus a push-button post-training pipeline that fine-tunes open-weight LLMs against an enterprise's private semantic model and data. Everything runs natively inside the customer's Snowflake account via Snowpark Container Services, including on Azure, so data, model, and reasoning stay in the customer's data cloud with no movement and inherit Snowflake's security and governance controls.
The company sells primarily to large enterprises already running on Snowflake, with a strategic focus on telecommunications (AT&T is the marquee customer and co-developer), banking and financial services (BNY), supply chain and manufacturing (Blue Yonder, Obeikan), and marketing. Revenue is metered in "Rel Units" billed hourly against the underlying Snowflake compute instance, packaged into four tiers: Standard ($2.00/Rel Unit), Enterprise ($3.00), Business Critical ($4.00 with Tri-Secret Secure and Tri-Secret keys), and Virtual Private (quote-based, fully isolated environment). Distribution combines self-serve install via the Snowflake Marketplace, direct enterprise sales for top-tier deployments, and a co-sell motion with Snowflake (2026 Telecom Product Partner of the Year) and AT&T. The company was founded by Molham Aref, employs roughly 166 people including 50+ PhDs, has cumulatively raised approximately $150M (with a $22.5M strategic investment from Snowflake Ventures and AT&T Ventures in December 2025), and reported $25.2M in ARR for 2024, up 41.6% from $17.8M in 2023.
RelationalAI firmographics
Firmographics- Name
- RelationalAI
- Legal name
- RelationalAI, Inc.
- Website
- https://relational.ai
- Company type
- Private
- Founded year
- 2017
- Operating status
- Operating
- Headcount range
- 101–250 employees
- Short description
- RelationalAI builds a Snowflake-native decision intelligence platform using its proprietary Rel declarative language, semantic modeling, and neuro-symbolic reasoners to run AI agents inside customer Snowflake accounts, primarily serving telecom, financial services, and supply chain enterprises.
- Ownership category
- akta.pro rank
RelationalAI industry classification
Industry- Product category
- Decision Intelligence Platform
- NAICS
- Custom Computer Programming Services (541511)
- SIC
- Services-Prepackaged Software (7372)
- akta.pro primary industry
- Enterprise AI Data & Knowledge Platforms (Vector Databases, Knowledge Graphs) (HDAEANAH)
- akta.pro secondary industry
- Knowledge Graphs, Ontologies & Entity Resolution (HDAAADAI)
Keywords
Where RelationalAI is headquartered
LocationHeadquarters
- HQ city
- Berkeley
- HQ country
- United States
- HQ region
- North America
Offices1 record
Markets served
RelationalAI business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Marketing or Sales, Infrastructure, Operations
Revenue model
- Rel Units consumption-based pricing: Primary revenue stream: usage-based pricing metered in "Rel Units" (a unit of reasoning analogous to compute hours). Rel Units are billed hourly based on the Snowflake compute instance type the customer chooses. Priced in tiered editions — Standard ($2.00/Rel Unit), Enterprise ($3.00/Rel Unit), Business Critical ($4.00/Rel Unit), and Virtual Private (quote-based, isolated environment). The Business Critical tier adds Tri-Secret Secure customer-managed keys; Virtual Private adds complete Snowflake environment isolation.
- Enterprise software ARR: RelationalAI generated $25.2M in revenue/ARR in 2024, a 41.6% increase from $17.8M in 2023 (GetLatka data). Land-and-expand motion through Snowflake Marketplace plus direct enterprise contracts.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | Standard — entry-level edition for starting projects |
| Usage-based | Pay-as-you-go | Enterprise — for high-growth and large-scale workloads |
| Usage-based | Pay-as-you-go | Business Critical — highest security and compliance demands |
| Other | Multi-year contract | Virtual Private — fully isolated Snowflake environment |
Go-to-market motion5 records
Distribution channels4 records
Marketing channels8 records
RelationalAI product offering
Product offeringCore offering
RelationalAI provides a Snowflake-native decision intelligence platform called Rel, combining a declarative programming language (Rel and its Python-native variant PyRel) for relational data with semantic modeling and four composable neuro-symbolic reasoners (rules, graph, predictive, prescriptive). The platform is delivered as the Knowledge Graph Coprocessor Snowflake Native App (running on Snowpark Container Services, including on Azure), allowing enterprises to build relational knowledge graphs and apply context-aware AI reasoning without moving data out of Snowflake.
Product overview
RelationalAI offers a single unified decision intelligence platform that runs natively inside the Snowflake AI Data Cloud, with the agentic decision agent "Rel" sitting on top. The core platform is delivered as the RelationalAI Knowledge Graph Coprocessor Snowflake Native App (GA November 2024, with Azure Snowpark Container Services GA February 2025), and is built on the Rel declarative programming language (a SIGMOD Research Highlight) and PyRel (its Python-native variant). On top of this core, RelationalAI offers modular capabilities announced and expanded at Snowflake Summit 2026: the Rel App (semantic modeling with natural language business queries), a prescriptive reasoner (GA), a predictive reasoner (graph neural networks), a graph reasoner, a rules-based reasoner, push-button post-training (private preview) for specializing open-weight LLMs, and RelationalAI coding agent skills that work with Claude Code, OpenAI Codex, GitHub Copilot, and Cortex Code. Consumption is metered in "Rel Units" across Standard, Enterprise, Business Critical, and Virtual Private editions. The full stack is positioned as a decision intelligence system extending Snowflake's AI Data Cloud with semantic models, advanced reasoners, and context-enhanced LLMs.
Differentiator
Problem solved
Functional benefit
Brands
- Rel: Rel is the company's flagship agentic decision intelligence system that runs natively inside the Snowflake AI Data Cloud. It is powered by the RelationalAI decision intelligence platform and uses semantic models, advanced reasoners, and context-enhanced LLMs to enable decision agents.
Products and services
- Rel
- RelationalAI Knowledge Graph Coprocessor (Snowflake Native App)
- Rel Programming Language
- PyRel
Quantifiable outcome
- Top of the leaderboard on the Spider 2.0 text-to-SQL benchmark (real-world database complexity)
- +3 more outcomes
Companies that use RelationalAI
Customer profileNamed customers7 records
Segments5 records
Ideal customer profiles4 records
RelationalAI technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration8 records
AI capability10 records
Feature7 records
RelationalAI partnerships and signals
Strategic signalPartnerships
15 partnerships are on record, tiered industry-body partner — relationalai is a launch partner for gsma open telco ai alongside at&t., integration partner — enables palantir ontology deployments to be ported into snowflake via osi and run advanced reasoning on relationalai without rebuild., tooling partner — coding-agent integration for rel/pyrel authoring., major si partner recognized at snowflake summit 2026 partner awards., major si partner recognized at snowflake summit 2026 partner awards. ey is also a relationalai customer (logo on home page)., supporting — contributed gpu compute for the joint at&t + relationalai 30 specialized telecom ai models., supporting — contributed gpu compute for the joint at&t + relationalai 30 specialized telecom ai models; broader strategic alignment via jensen huang's public statements on post-training open-weight models., flagship — joint co-development, joint benchmarks, and marquee customer reference., flagship ecosystem initiative — relationalai co-founded osi with snowflake and industry leaders and leads the osi ontology representation working group., co-marketing partner at mobile world congress 2025. and flagship — deepest strategic alliance; rel runs natively as a snowflake native app inside the snowflake ai data cloud, with co-marketing (snowflake 2026 telecom product partner of the year), co-engineering (snowpark container services, cowork integration, snowflake intelligence), and joint go-to-market..
- GSMA (Open Telco AI)industry-body partner — relationalai is a launch partner for gsma open telco ai alongside at&t.RelationalAI collaborated with AT&T to deliver 33 best-in-telco LLMs (OTel) as part of the GSMA Open Telco AI launch.
- Palantirintegration partner — enables palantir ontology deployments to be ported into snowflake via osi and run advanced reasoning on relationalai without rebuild.As a launch partner in the OSI initiative, RelationalAI enables enterprises with existing Palantir ontology deployments to port semantic models into Snowflake via OSI and run advanced reasoning on RelationalAI with no rebuild required.
- Anthropic (Claude Code)tooling partner — coding-agent integration for rel/pyrel authoring.RelationalAI integrated its coding agent skills with Claude Code, plus OpenAI Codex and GitHub Copilot, so developers can extend RelationalAI models directly from their preferred coding environments.
- OpenAI (Codex)tooling partner — coding-agent integration for rel/pyrel authoring.RelationalAI integrated its coding agent skills with OpenAI Codex, Claude Code, and GitHub Copilot so developers can extend RelationalAI models from their preferred coding environments.
- GitHub (Copilot)tooling partner — coding-agent integration for rel/pyrel authoring.RelationalAI integrated its coding agent skills with GitHub Copilot, Claude Code, and OpenAI Codex so developers can extend RelationalAI models from their preferred coding environments.
- Deloittemajor si partner recognized at snowflake summit 2026 partner awards.Deloitte was among the top Snowflake partners recognized at the 2026 Snowflake Partner Awards for technical innovation and business results.
- Accenturemajor si partner recognized at snowflake summit 2026 partner awards.Accenture was among the top Snowflake partners recognized at the 2026 Snowflake Partner Awards for technical innovation and business results.
- EYmajor si partner recognized at snowflake summit 2026 partner awards. ey is also a relationalai customer (logo on home page).EY was among the top Snowflake partners recognized at the 2026 Snowflake Partner Awards for technical innovation and business results.
- AMDsupporting — contributed gpu compute for the joint at&t + relationalai 30 specialized telecom ai models.AMD provided GPU contributions supporting the 30 specialized telecom AI models released jointly by AT&T and RelationalAI in March 2026.
- Nvidiasupporting — contributed gpu compute for the joint at&t + relationalai 30 specialized telecom ai models; broader strategic alignment via jensen huang's public statements on post-training open-weight models.Nvidia provided GPU contributions supporting the 30 specialized telecom AI models released jointly by AT&T and RelationalAI in March 2026; Jensen Huang cited post-training of open-weight models and structured data fusion as "the foundation of trustworthy AI" at GTC keynote.
- AT&Tflagship — joint co-development, joint benchmarks, and marquee customer reference.AT&T and RelationalAI jointly contribute to GSMA Open Telco AI and topped Generative AI leaderboards (Spider, TeleLogs). They have collaborated on multiple AI solutions, with AT&T deploying the technology throughout its operations. In March 2026 they announced 30 specialized telecom AI models trained on industry standards with GPU contributions from AMD and Nvidia.
- Open Semantic Interchange (OSI) industry coalitionflagship ecosystem initiative — relationalai co-founded osi with snowflake and industry leaders and leads the osi ontology representation working group.RelationalAI joined Snowflake and industry leaders to establish the Open Semantic Interchange (OSI), an open, vendor-neutral standard for expressing business semantics; the initiative extended OSI to ontological semantics in May 2026.
- Ericssonco-marketing partner at mobile world congress 2025.Partnered with Snowflake at Mobile World Congress 2025 in Barcelona (Booth 5A31 in Hall 5) alongside Ericsson and Nokia, showcasing AI Data Cloud platform for telecom operators.
- Nokiaco-marketing partner at mobile world congress 2025.Partnered with Snowflake at Mobile World Congress 2025 in Barcelona (Booth 5A31 in Hall 5) alongside Nokia and Ericsson, showcasing AI Data Cloud platform for telecom operators.
- Snowflakeflagship — deepest strategic alliance; rel runs natively as a snowflake native app inside the snowflake ai data cloud, with co-marketing (snowflake 2026 telecom product partner of the year), co-engineering (snowpark container services, cowork integration, snowflake intelligence), and joint go-to-market.Rel is delivered as a Snowflake Native App (listing GZTYZOOIX8H) running on Snowpark Container Services, generally available since November 12, 2024, and on Azure since February 3, 2025. New Rel App, prescriptive and predictive reasoners, and integration with Snowflake CoWork were announced at Snowflake Summit 2026. Together they launched and lead the Open Semantic Interchange (OSI) industry initiative.
Scale indicators11 records
Recent moves6 records
Expansion highlights6 records
RelationalAI competitors and assessment
Company assessmentDirect peers
- Palantir Technologies: Palantir's Foundry and AIP platforms deliver ontology-driven decision intelligence for enterprises and governments. Directly comparable to RelationalAI's Rel platform because both rely on semantic models + reasoners for grounded decision-making; Palantir is also an OSI launch partner enabling ontology portability into RelationalAI/Snowflake.
- Neo4j: Neo4j is the leading dedicated graph database and knowledge-graph platform, with a growing focus on GraphRAG and GenAI-driven decision intelligence. Most directly comparable to RelationalAI's graph reasoner and relational knowledge graph approach, though delivered as a standalone graph store rather than a Snowflake-native app.
- TigerGraph: TigerGraph provides an enterprise-scale native graph database with analytics and machine-learning capabilities. Comparable to RelationalAI's graph reasoner and PathFinder engine for multi-hop graph traversal, fraud detection, and supply-chain use cases; targets the same regulated enterprise buyers.
- Stardog: Stardog is an enterprise knowledge-graph platform that combines graph database, semantic modeling, and AI reasoning. Directly comparable to RelationalAI's ontology-driven decision intelligence for BFSI, life sciences, and government customers; competes for the same semantic-layer budget.
Broad incumbents
- Databricks: Databricks offers Mosaic AI and a unified data intelligence platform that competes in the same AI-on-structured-data category. It is the principal horizontal alternative to Snowflake + RelationalAI for enterprises building knowledge-graph and decision-intelligence workloads on a data lakehouse.
- DataRobot: DataRobot provides an enterprise AI platform for predictive and prescriptive modeling with strong governance. Overlaps with RelationalAI's predictive (GNN) and prescriptive reasoners; targets the same Fortune 500 buyers building decision-intelligence use cases on structured enterprise data.
- H2O.ai: H2O.ai offers an open-source-driven enterprise AI platform spanning AutoML, LLM fine-tuning, and decision intelligence. Comparable to RelationalAI's predictive reasoner and push-button LLM superalignment, especially in regulated industries; competes for similar data-science and AI-platform budgets.
- Microsoft (Microsoft Fabric + Synapse): Microsoft Fabric unifies data engineering, analytics, and AI on OneLake with deep Azure integration. Comparable to RelationalAI as an alternative data-cloud + decision-intelligence stack; Azure GA of RelationalAI's Snowflake Native App makes Microsoft an important adjacent (and competitive) hyperscaler.
Emerging players
- Domino Data Lab: Domino provides an enterprise MLOps and decision-intelligence platform with model governance and knowledge-graph capabilities. Comparable to RelationalAI's predictive reasoner + Snowflake-native deployment pattern; targets the same large-enterprise data-science teams.
- Anaconda: Anaconda's AI platform provides data-science tooling and Python-native data infrastructure for enterprises. Adjacent to RelationalAI's PyRel/Python strategy; comparable in the developer-led bottom-up adoption motion and Snowflake ecosystem presence.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks7 records
Key highlights7 records
Customer concentration
RelationalAI social profiles
Digital presenceRelationalAI financial estimates
Financial estimateRevenue estimate
Valuation estimate
RelationalAI leadership team
Management profileNumber of profiles
Profiles6 records
RelationalAI funding detail
Funding detailFunding overview
Funding rounds3 records
Investors6 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
RelationalAI M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about RelationalAI
What does RelationalAI do?
RelationalAI provides a Snowflake-native decision intelligence platform called Rel, combining a declarative programming language (Rel and its Python-native variant PyRel) for relational data with semantic modeling and four composable neuro-symbolic reasoners (rules, graph, predictive, prescriptive). The platform is delivered as the Knowledge Graph Coprocessor Snowflake Native App (running on Snowpark Container Services, including on Azure), allowing enterprises to build relational knowledge graphs and apply context-aware AI reasoning without moving data out of Snowflake.
Is RelationalAI a public or private company?
RelationalAI is a private company. It is classified as venture growth investor backed and is currently operating.
When was RelationalAI founded?
RelationalAI was founded in 2017. It employs 101 to 250 people.
Where is RelationalAI based?
RelationalAI is headquartered in Berkeley, United States, in the North America region.
How does RelationalAI make money?
Two revenue lines are on record. Rel Units consumption-based pricing is the primary driver. The others are enterprise software ARR.
Who are RelationalAI's main competitors?
Direct peers on record are Palantir Technologies, Neo4j, TigerGraph and Stardog. Broad incumbents are Databricks, DataRobot, H2O.ai and Microsoft (Microsoft Fabric + Synapse). Emerging players are Domino Data Lab and Anaconda.
Does RelationalAI have an API?
Yes. RelationalAI offers the Rel API for building decision agents, along with PyRel, its Python-native language for building, querying, and reasoning over ontologies. Developers can start building with the Rel API or install the native app inside Snowflake. The platform exposes a Semantic Model API, natural language reasoning, rules-based reasoning, graph reasoning, prescriptive reasoning, and predictive reasoning. Documentation is hosted at docs.relational.ai. Developer documentation is at docs.relational.ai.
What industry is RelationalAI in?
RelationalAI's product category is Decision Intelligence Platform. Its primary akta.pro industry code is HDAEANAH, Enterprise AI Data & Knowledge Platforms (Vector Databases, Knowledge Graphs), with a secondary code of HDAAADAI, Knowledge Graphs, Ontologies & Entity Resolution. Its NAICS code is 541511 and its SIC code is 7372.