HeliosDB Full · C-Level Enterprise Unification (Rung 4)

14 Native Wire Protocols. Zero-ETL Lakehouse.
Cut Multi-Vendor Data Estate TCO by 83%.

HeliosDB Full unifies transactional databases (PostgreSQL, MySQL, Oracle 23ai with 20 DBMS_* packages, SQL Server, IBM Db2), caching and queues (Redis with Lua 5.1 & BullMQ), wide-column and document stores (Cassandra, MongoDB), analytical warehouses (ClickHouse, Snowflake, Databricks Iceberg/Delta), and vector search (Pinecone, REST, GraphQL) in one self-healing engine.

14Native Protocols Across OLTP, Cache, Warehouse, Lakehouse & Vector
50-100xMulti-Tenant SaaS Density via Adaptive Cold-to-Hot Hibernation (<50ms Wake)
0 PagesAutonomous Self-Healing Storage (Bayesian Root-Cause + RL Repair)
PQC + FHENIST Kyber-768 / Dilithium-3 Post-Quantum TLS & CKKS Homomorphic Encryption

Test-Drive HeliosDB Full's Deep-Tech Differentiators

Explore how Autonomous Self-Healing Storage, CKKS Homomorphic Encryption, 50-100x Tenant Hibernation, 14-Protocol Wire Multiplexing, and Merkle Audit Lineage work interactively.

HeliosDB Full

The HeliosDB edition for enterprise workloads. HeliosDB Full serves the PostgreSQL wire protocol and an HTTP gateway on the shared HeliosDB SQL core, with opt-in MySQL, Oracle and Redis listeners (MongoDB, Cassandra and SQL Server compatibility is planned); the gateway serves vector stores with hybrid search, embeddings, chat completions and agent memory. In-database model training and quantum-inspired optimization are planned. HeliosDB Full runs as a single server today: Raft clustering, multi-region replication and sharding are planned, and GPU acceleration (NVIDIA CUDA) is available in HeliosDB Full builds on request and is not enabled in the public release build.

~200 MB Binary (internal build)
PG + HTTP On by default; other listeners opt-in
26 Use Cases
SQL
-- Multi-protocol: PostgreSQL, MySQL,
-- Oracle, Redis + HTTP;
-- Mongo, Cassandra, SQL Server: planned

-- PostgreSQL, MySQL, Oracle and HTTP
-- read and write the same tables
SELECT region, SUM(revenue)
FROM sales
GROUP BY region;
Agent-Ready

Drop Full into Claude Code & Codex CLI in 30 seconds

Ships an AGENTS.md aggregate plus a .claude/skills/ catalogue. Every CLI flag, REPL meta-command and public API indexed for LLM coding agents. Read the guide →

# source provided during onboarding
cargo install --path ./heliosdb-full
bash scripts/install-skills.sh

Which AI retrieval features does HeliosDB Full offer?

HeliosDB Full offers hybrid search through its HTTP vector-store API. In-database machine learning, graph traversal and hybrid search as SQL functions, Graph-RAG, Code-Graph and a native MCP server are planned for Full; Code-Graph, Graph-RAG and the MCP server ship in HeliosDB Nano today.

Hybrid search

Over the HTTP vector-store API

Combine full-text and vector search with POST /api/v1/vectors/stores/{store}/search/hybrid on the HTTP gateway.

Graph in SQL · planned

graph_traverse()

Graph traversal and shortest-path search as SQL table functions, and hybrid search as a SQL function, are planned.

Native MCP · planned

MCP server for AI agents

A native MCP server for HeliosDB Full is planned. HeliosDB Nano ships one today as a build-from-source feature.

In-database ML · planned

Train where the data lives

Model training in SQL, anomaly detection, time-series forecasting and a model registry are planned for HeliosDB Full. The HTTP gateway serves embeddings, chat completions and agent memory today.

Edition comparison

How is HeliosDB Full different from Lite?

Full adds MySQL, Oracle and Redis listeners next to PostgreSQL and an HTTP gateway with vector stores, embeddings and agent memory; MongoDB, Cassandra and SQL Server compatibility is planned. HeliosDB Full runs as a single server today; Raft clustering, multi-region replication and sharding are planned, and GPU acceleration (NVIDIA CUDA) is available in HeliosDB Full builds on request and is not enabled in the public release build. Nano and Lite run time-travel queries (AS OF) and Git-like branches; for HeliosDB Full both are planned (the current release accepts the syntax but does not read historical snapshots or create branches), and so are in-database model training, SSO, LDAP and attribute-based access control (edition comparison).

Which database protocols can connect to HeliosDB Full?

HeliosDB Full serves PostgreSQL, MySQL, Oracle and Redis clients plus an HTTP gateway. The PostgreSQL listener and the HTTP gateway are on by default; the others are switched on with a startup flag (protocol hub). Each database engine below carries its status in the Release and Capability Matrix: Available, Available with limitations, In preview (included in the edition and usable for evaluation, but not yet verified by HeliosDB on any version, and not supported) or Planned. Each status is the best result that HeliosDB's tests or its own production use show on any version or build of HeliosDB Full, so it names no release and does not mean that every version was tested (status summary); the statuses come from a draft of the matrix, which is informational until HeliosDB publishes a frozen version (Release and Capability Matrix terms, clause 7.1). Driver compatibility does not imply complete feature parity with the original database engine, or compatibility with every application, extension or administrative tool.

  • PostgreSQL (5432), Available with limitations: the primary interface for drivers, ORMs and tools; feature parity with PostgreSQL is not claimed. HeliosDB Full listeners do not terminate TLS yet; put a TLS-terminating proxy in front, and native PostgreSQL TLS (with post-quantum key exchange preferred) is planned for the next release
  • MySQL (3306), Available with limitations: transaction control through this listener is not covered (treat each statement as committing on its own), and MySQL-only syntax may need changes
  • Oracle (1521), Available with limitations: tested with python-oracledb in thin mode for core DDL, DML and common queries; every statement commits as it runs (driver commit and rollback calls are not supported), Oracle-specific syntax and PL/SQL need rewriting, and the logon password must be set explicitly
  • Redis (6379), In preview: RESP2 and a subset of the Redis commands, usable for evaluation but not supported; the Transaction Durability Commitment does not cover it, because it excludes in-memory key-value stores
  • MongoDB, Cassandra and SQL Server, Planned: reading and writing HeliosDB tables through these protocols is on the roadmap, and so are GridFS and capped collections for MongoDB; use the PostgreSQL interface today
  • The MySQL and Redis listeners do not verify passwords yet, so keep them on a trusted network
  • Planned beyond the database engines: an S3-compatible object API. Reach the data over PostgreSQL or the HTTP gateway (POST /api/v1/query) today
Bash
# PostgreSQL and the HTTP gateway are on by default;
# the other listeners are opt-in at startup
# (listeners bind 127.0.0.1 unless told otherwise)
heliosdb-full --data-dir /var/lib/heliosdb \
  --bind-addr 10.0.0.12 --allow-external \
  --enable-mysql --enable-oracle --enable-redis

# Connect via PostgreSQL
psql -h 10.0.0.12 -p 5432 -d mydb

# Connect via MySQL
mysql -h 10.0.0.12 -P 3306 -u postgres

# Connect via Redis (AUTH with a role name;
# the password is not verified yet)
redis-cli -h 10.0.0.12 -p 6379 --user app --pass unused

# PostgreSQL and MySQL read and write the same tables

Does HeliosDB Full replicate data across regions?

Not in the current release. HeliosDB Full runs as a single server today; Raft clustering, multi-region replication and sharding are planned.

Can you train machine-learning models inside HeliosDB Full?

Not in the current release. HeliosDB Full serves vector stores with hybrid search, embeddings, chat completions and agent memory over its HTTP gateway; in-database model training in SQL, the neural query planner, anomaly detection and forecasting, NL-to-SQL, quantum-inspired optimization, and SQL vector indexes with Product Quantization are planned.

Quantum-Inspired Optimization (planned)

Quantum-inspired query optimization (QAOA, VQE and Grover-inspired search) is a research item for HeliosDB Full. It is planned and not part of the current release build.

Does HeliosDB Full support GPU acceleration?

GPU acceleration (NVIDIA CUDA) is available in HeliosDB Full builds on request and is not enabled in the public release build; contact . The CUDA work covers hash joins, group-by aggregation, sorting and vector similarity search. AMD ROCm support is planned.

Neuromorphic Computing (planned)

Pattern recognition and anomaly detection with spiking neural networks on Intel Loihi 2 neuromorphic processors is a research item for HeliosDB Full. It is planned and not part of the current release build.

How does HeliosDB Full control access to data?

HeliosDB Full authenticates PostgreSQL clients with SCRAM-SHA-256; SQL GRANT/REVOKE enforcement, row-level security, encryption at rest with key rotation and cloud KMS, ABAC, SAML/OIDC single sign-on and LDAP are planned. A compliance framework for SOC 2, HIPAA, GDPR and PCI-DSS, blockchain-backed audit trails and post-quantum cryptography are planned too. HeliosDB Full listeners do not terminate TLS yet; put a TLS-terminating proxy in front, and native PostgreSQL TLS (with post-quantum key exchange preferred) is planned for the next release.

Can you build a custom HeliosDB Full binary?

Yes. Licensed customers build HeliosDB Full from source with compile-time features (73 in the current server), leaving out components they do not use or adding research features. Custom binary sizes range from about 60 MB to 200 MB (internal builds, approx.).

Research Features (planned)

Genomics sequence storage and alignment, blockchain-CRDT synchronization, holographic 3D visualization and multimodal storage for text, images, audio and video are research items for HeliosDB Full. They are planned and not part of the current release build. Graph-RAG is planned for Full too; it ships in HeliosDB Nano today.

Frequently Asked Questions

What is HeliosDB Full?

HeliosDB Full is an enterprise database server that serves several database wire protocols, plus an HTTP gateway with vector stores, embeddings and agent memory, from one engine. HeliosDB Full runs as a single server today; Raft clustering, multi-region replication and sharding are planned, and so are in-database model training and enterprise access control (SSO, LDAP, ABAC). It is licensed SSPL-1.0 and provided to licensed customers during onboarding (edition comparison).

Which database clients can connect to HeliosDB Full?

One HeliosDB Full server serves PostgreSQL, MySQL, Oracle and Redis clients plus an HTTP gateway; PostgreSQL and the HTTP gateway are on by default and the others are enabled with startup flags. In the Release and Capability Matrix, where each status is the best result on any version of HeliosDB Full, PostgreSQL, MySQL and Oracle are Available with limitations, Redis is In preview, and MongoDB, Cassandra and SQL Server are Planned. The PostgreSQL, MySQL and Oracle listeners and the HTTP gateway read and write the shared tables (protocol hub).

How should applications connect to HeliosDB Full?

Use a PostgreSQL driver such as psycopg, pgx, node-postgres, JDBC or libpq: the PostgreSQL wire protocol is the primary interface. HeliosDB Full listeners do not terminate TLS yet, so put a TLS-terminating proxy in front; native PostgreSQL TLS is planned for the next release. Full has no gRPC client API, and the REST clients in the HeliosDB SDKs target Nano and Lite, not Full (FAQ).

Can a MongoDB application move to HeliosDB Full unchanged?

No, not yet. MongoDB compatibility for HeliosDB Full is Planned in the Release and Capability Matrix: reading and writing HeliosDB tables over the MongoDB protocol, GridFS and capped collections are on the roadmap. Use the PostgreSQL interface for data access today (compatibility notes).

Are writes in HeliosDB Full durable when acknowledged?

Yes, by default. A write is fsync-durable when the client receives the acknowledgement, and concurrent writers share one flush through group commit. An opt-in fast mode acknowledges before the flush and bounds the possible loss to the write-ahead log flush interval (release notes).

Does HeliosDB Full need HeliosProxy?

No. HeliosDB Full runs as a single server today and serves clients directly; its own clustering and failover are planned. You can put HeliosProxy, the separate Apache 2.0 proxy, in front of Full when you want its pooling, failover, policy or observability features.

How is HeliosDB Full licensed and priced?

HeliosDB Full is source-available under SSPL-1.0, with a commercial license for terms other than SSPL. Pricing is custom on an annual contract through , and binaries, container images and Helm charts are provided during onboarding rather than through public registries (pricing).

How does HeliosDB Full control access?

HeliosDB Full authenticates PostgreSQL clients with SCRAM-SHA-256; SQL GRANT/REVOKE enforcement, row-level security, encryption at rest with key rotation and cloud KMS, ABAC, SAML/OIDC single sign-on and LDAP are planned. HeliosDB Full listeners do not terminate TLS yet, and native PostgreSQL TLS is planned for the next release. The MySQL and Redis listeners do not verify passwords yet.

Who should use HeliosDB Full?

Use HeliosDB Full when one licensed server must serve several database protocols; enterprise access control (SSO, LDAP, ABAC), multi-region replication and Raft clustering are planned.

Enterprise & Fortune 500

PostgreSQL, MySQL, Oracle and Redis clients plus an HTTP gateway in one server. ABAC, SSO, LDAP and a SOC 2/HIPAA/GDPR compliance framework are planned.

Multi-Region Deployments (planned)

Raft clustering across regions with configurable consistency levels, automatic failover and geo-aware routing is planned; HeliosDB Full runs as a single server today.

Regulated Industries (planned)

Compliance automation, blockchain-backed audit trails, data-sovereignty controls and post-quantum cryptography for healthcare, finance and government are planned for HeliosDB Full.

Research & Innovation

GPU acceleration (NVIDIA CUDA) in HeliosDB Full builds on request; quantum-inspired optimization, genomics, holographic visualization and neuromorphic computing are research items on the roadmap.

Ready for the Frontier?

Contact our sales team to discuss enterprise licensing and deployment for HeliosDB Full.