H

Hasura

Indian-founded GraphQL data-API engine and AI data-agent platform — launched 2017 by Tanmai Gopal (CEO) and Rajoshi Ghosh from dual Bangalore + San Francisco headquarters; $136.5M total funding including a $100M Series C on 22 February 2022 at a $1 billion unicorn valuation led by Greenoaks with Nexus Venture Partners, Lightspeed Venture Partners and Vertex Ventures; 2024-2026 strategic pivot to Hasura DDN (Data Delivery Network) + PromptQL (enterprise AI data agents with seven-figure customer deals)

GraphQL Engine / Data API / AI Data Agents 4.6 / 5 (1 Rating) Hasura OSS / Cloud Free / DDN active-model pricing / PromptQL prepaid consumption Updated May 2026 🇮🇳 Indian-founded; Bangalore engineering centre; self-host OSS available
✅ Recommended for Indian engineering teams — strongest Indian-founded data-API platform with credible 2024-2026 AI pivot

Quick Verdict

Hasura is one of the most successful Indian-founded developer-tools companies on the global stage — built originally around a viral open-source GraphQL engine that instantly generates production-ready APIs from Postgres / MySQL / SQL Server / BigQuery / Snowflake databases, and pivoted in 2024-2026 into an AI data-agent platform that's now closing seven-figure enterprise contracts. The company was founded in 2017 by Tanmai Gopal (CEO) and Rajoshi Ghosh with dual headquarters in Bangalore, India and San Francisco, California — making Hasura one of the cleanest examples of an Indian-founded developer-infrastructure company built deliberately for the global market from day one. Across four funding rounds Hasura has raised approximately $136.5 million in total, including the headline $100 million Series C on 22 February 2022 at a $1 billion unicorn valuation — led by Greenoaks Capital with returning investors Nexus Venture Partners, Lightspeed Venture Partners, Vertex Ventures and STRIVE all participating. The original Hasura GraphQL Engine is still actively maintained as open-source under the Apache 2.0 licence and remains widely deployed by Indian engineering teams who want instant GraphQL APIs without writing backend code. The biggest strategic shift in the company's history is the 2024-2026 pivot to Hasura DDN (Data Delivery Network) and PromptQL — Hasura is no longer pitching itself only as a GraphQL data-API layer, it's now pitching itself as a "100% reliable enterprise AI data-agent platform". PromptQL has been deployed at large financial-services organisations with deployments reaching 25,000 users at single accounts, and Hasura has built a notable consulting model around $900/hour AI engineers as a premium-services layer on top of the platform. For Indian engineering teams the right framing is: Hasura is the right call for teams building GraphQL APIs on existing databases (the original product still works beautifully), and increasingly for AI-platform / fintech / BFSI teams deploying production AI agents that need to reliably query structured data. The wrong call for greenfield BaaS-style mobile apps (use Appwrite / Supabase) or for teams that don't actually want GraphQL.

Indian-founder pedigree (Tanmai Gopal)
4.8
GraphQL engine quality (original product)
4.7
PromptQL / AI data-agent traction (2024-2026)
4.4
Self-host OSS for RBI / BFSI
4.6
Greenfield BaaS-style mobile-app fit
2.7

What is Hasura?

Hasura was originally — and still primarily, in most production deployments — a developer tool that connects to your existing database (Postgres, MySQL, SQL Server, BigQuery, Snowflake, MongoDB and many more) and instantly composes a production-ready GraphQL API on top of the schema. The mental model: point Hasura at a Postgres database, and within minutes you have a secure, scalable GraphQL endpoint with row-level authorization, subscriptions, mutations, query plans and connection pooling — without writing a single line of backend resolver code. Any subsequent schema change (adding a column, creating a table, modifying a relationship) is instantly reflected in the GraphQL API without a deploy. The product surface has expanded to also cover REST API generation, event triggers (database events → external webhooks / serverless functions), scheduled triggers (cron-style), actions (custom GraphQL operations backed by your own business-logic code), and remote schemas (federate multiple GraphQL services into one schema).

The company was founded in 2017 in Bangalore, India by Tanmai Gopal (CEO) and Rajoshi Ghosh. Tanmai had previously co-founded a consulting firm (34Cross Systems) and was a long-time contributor in the Indian developer-tools community. The bet was that the most common bottleneck in modern application development was the time engineers spent writing boilerplate CRUD APIs over their databases, and that a code-free GraphQL engine that auto-generated those APIs (with proper authorization, performance, and operations tooling baked in) could collapse weeks of backend work into minutes. The bet paid off: Hasura's open-source GraphQL Engine went viral on GitHub, became one of the most-starred OSS data-API projects, and was deployed at scale by enterprises and startups globally.

Funding history: $1.6M Seed in 2018 → $9.9M Series A in 2020 (led by Nexus Venture Partners) → $25M Series B in 2021 (led by Lightspeed) → $100M Series C on 22 February 2022 led by Greenoaks Capital with Nexus / Lightspeed / Vertex / STRIVE participating, valuing the company at $1 billion and making Hasura one of the early Indian-founded developer-tools unicorns. Total funding to date is approximately $136.5 million. Headquarters are dual: Bangalore (where most of the engineering team is based) and San Francisco, California (where the US go-to-market and enterprise sales operate). This structure — Indian engineering depth + US go-to-market — is increasingly common among Indian-founded developer-infra companies (Postman, Browserstack, Freshworks, Hasura, Razorpay's Curlec) and has become a template for Indian SaaS founders building for global markets.

The single most important strategic event in the company's recent history is the 2024-2026 pivot to AI data agents. In response to the post-ChatGPT shift in what enterprises want from data infrastructure, Hasura launched two products that materially expanded the company's positioning: (1) Hasura DDN (Data Delivery Network) in 2024 — a redesigned data API platform with active-model-based pricing (you pay per data model you actively use, defined as ≥1,000 hits/month, rather than per query or per seat); and (2) PromptQL — an enterprise AI data-agent platform that lets engineers deploy AI agents that can analyze internal company data using LLMs with guaranteed reliability. PromptQL has been deployed at large financial-services organisations with single-account deployments reaching 25,000 users across multiple subsidiaries; Hasura has also built a premium consulting layer around $900/hour AI engineers tasked with custom PromptQL implementations at enterprise customers, generating seven-figure annual contracts.

What Hasura gives you (the 2026 product surface)

⚡ Instant GraphQL on any database

Point Hasura at Postgres / MySQL / SQL Server / BigQuery / Snowflake / MongoDB / Oracle / Citus / YugabyteDB / CockroachDB and get a production-ready GraphQL API in minutes. Schema changes propagate instantly. The original product surface that built the company.

🔐 Granular authorization (row-level)

Configure rules like "users can only query their own orders" or "admins can read all but only write within their tenant" at the database layer. Hasura's authorization model is genuinely category-leading and is the standout feature for fintech / BFSI multi-tenancy.

📡 Event triggers + scheduled triggers + actions

Subscribe to database events (row inserted, updated, deleted) and trigger external webhooks. Run cron-style scheduled triggers. Layer custom business logic via Actions backed by your own serverless functions. Makes async architectures dramatically simpler.

🌐 Hasura DDN (Data Delivery Network, 2024)

Redesigned data API platform built for federation across multiple data sources with active-model-based pricing (you pay per model that's actively used, threshold 1,000 hits/month). Aligns cost with value rather than per-query or per-seat.

🤖 PromptQL (AI data agents, 2024-2026)

Enterprise AI data-agent platform — engineers deploy LLM-backed agents that analyze internal company data with guaranteed reliability. Major Indian / global financial-services customer base; single-account deployments reaching 25,000 users.

🏠 Open-source Hasura GraphQL Engine

Apache 2.0-licensed, self-hostable. Production-grade and used at meaningful scale by paying customers' OSS deployments. Critical for Indian BFSI / NBFC / fintech teams under RBI / data-residency rules — run on your own infrastructure, never send data to Hasura Cloud.

Pricing & plans (2026)

Hasura has restructured pricing several times since launching DDN and PromptQL. Live rates from hasura.io/pricing:

  • Hasura GraphQL Engine OSSfree forever (Apache 2.0). Self-host on Docker / Kubernetes / any cloud. Production-grade. Best for Indian developers and small startups, RBI-strict BFSI / fintech, and any team with DevOps maturity.
  • Hasura Cloud Free — generous free tier on managed Hasura Cloud, useful for prototyping and small projects. Limits on data-pass-through, requests per minute and connection counts.
  • Hasura DDNactive-model-based pricing: you pay per active data model (defined as ≥1,000 hits/month). Pricing scales by model count rather than per-query or per-seat. Most-bought tier for Indian Series A+ SaaS teams running federated data APIs.
  • Hasura Enterprise — custom pricing with SSO (SAML / OIDC), advanced RBAC, audit logs, on-prem deployment option, dedicated CSM. Typical Indian BFSI / large-enterprise contracts $50K-$500K/year.
  • PromptQLprepaid consumption pricing. Customers purchase a set amount of consumption dollars; usage deducts from the balance. Enterprise deployments cluster at seven-figure annual contracts where PromptQL is the platform for 1,000+ active users analysing internal company data via LLMs.
  • PromptQL AI Engineers (services) — premium consulting layer at approximately $900/hour for custom AI-agent implementations. Used by large enterprises for production-critical AI agent rollouts.

For Indian buyers, all Cloud pricing is in USD with 18% IGST applicable. The OSS self-hosted engine running on AWS Mumbai or Hetzner for a typical Indian Series A SaaS team costs roughly ₹15,000-₹50,000/month all-in (compute + DB + DevOps engineer time). Hasura Cloud DDN typical Indian SaaS spend clusters at $200-$2,000/month depending on model count and traffic. Indian fintech / BFSI Enterprise contracts (with on-prem option, SSO, audit logs, dedicated CSM) typically run ₹40 lakh – ₹4 crore/year.

When Hasura is the right call

  1. You have an existing Postgres / MySQL / Snowflake / BigQuery database and need a fast GraphQL API on top — Hasura's original product is still the cleanest way to do this. Indian Series A+ SaaS teams building data-heavy apps (analytics dashboards, internal tools, customer portals) get a working GraphQL API in hours instead of weeks of backend dev.
  2. You're an Indian BFSI / fintech / NBFC needing self-host + row-level authorization at scale — Hasura's open-source engine with its granular permission model is genuinely category-leading for multi-tenant fintech apps. RBI / data-residency compliance through self-host on AWS Mumbai or on-prem.
  3. You're an Indian enterprise deploying AI agents that need to reliably query internal data — PromptQL is the 2024-2026 strategic story. Most credible "agentic AI with structured-data reliability" platform on the market, with proven seven-figure enterprise deployments.
  4. You want to support an Indian-founded developer-infra company building globally — Tanmai Gopal and Rajoshi Ghosh are some of the most respected Indian developer-infra founders. Bangalore engineering depth, US go-to-market, with serious investor backing (Greenoaks, Lightspeed, Nexus, Vertex).
  5. You're federating data across multiple sources — Hasura DDN's federation model lets you build one GraphQL API that pulls from Postgres + Snowflake + BigQuery + REST endpoints + MongoDB simultaneously. Useful for Indian data-engineering teams running multi-cloud architectures.

Hasura is the wrong call when: you want a full BaaS for mobile / web apps with auth / storage / push notifications baked in (use Appwrite or Supabase — Hasura is a data-API layer, not a full BaaS); you specifically don't want GraphQL and prefer REST or gRPC (PostgREST and gRPC frameworks are better fits); you're a single-developer side project that hasn't reached production data complexity (just use Prisma / Drizzle directly); or you need full SQL-first Postgres developer experience with edge functions and built-in auth (Supabase is more aligned there).

Pros & cons

✓ Pros

  • Indian-founded global developer-infra company — Tanmai Gopal + Rajoshi Ghosh credibility
  • Apache 2.0 OSS engine — fully self-hostable, RBI / data-residency friendly
  • Instant GraphQL from any database — original product still category-leading
  • Row-level authorization is best-in-class for multi-tenant fintech / BFSI apps
  • $136.5M raised at $1B unicorn valuation; Greenoaks / Lightspeed / Nexus / Vertex backing
  • Dual Bangalore + SF HQ — IST-aligned support and Indian engineering depth
  • PromptQL is a credible 2024-2026 AI data-agent story with seven-figure enterprise deals
  • Hasura DDN active-model pricing aligns cost with value (vs per-query / per-seat)
  • Active maintenance pace; large global developer community

✗ Cons

  • Not a full BaaS — no built-in auth / storage / push notifications (Appwrite / Supabase better there)
  • Cloud pricing in USD with 18% IGST; no INR billing option
  • GraphQL has fallen out of favour in some 2024-2026 engineering circles vs tRPC / REST
  • Strategic shift to PromptQL has reduced focus on the OSS GraphQL engine community
  • DDN is the recommended path forward but adds complexity vs the original Hasura engine
  • $900/hour AI Engineers consulting model is premium-priced; not for budget-conscious buyers
  • Documentation for DDN + PromptQL still maturing vs the original engine docs
  • Self-hosted operations require real DevOps maturity at scale

Related insights & playbooks