/Docs
Docs›Get Started›Product›Features

Features

Explore Tacnode's capabilities — unified storage, real-time indexing, vector search, and more.

Unified Context, Every Workload

Tacnode Context Lake maintains shared, live, and semantic context in a single system. This page covers the technical capabilities that make this possible.

Distributed Relational Database

Tacnode is a distributed relational database with full ACID transactions.

Single-machine relational databases face performance bottlenecks that limit their ability to meet business demands. Tacnode enables horizontal scaling — add machines and resources to increase throughput. Distributed transaction support ensures strong data consistency during expansion.

Key capabilities:

  • Horizontal scaling with distributed transactions
  • Serializable isolation for strong consistency
  • JSON type for semi-structured data
  • Vector type for embeddings and AI workloads

Flexible Data Modeling

Tacnode handles structured, semi-structured, and vector data within the same transactional system.

Structured data: Traditional tables with schemas, foreign keys, and referential integrity.

Semi-structured data: The JSON type handles schema-flexible documents. Unlike document databases that struggle with many-to-many relationships, Tacnode combines JSON flexibility with relational joins and foreign keys — giving you the best of both models.

Vector data: Native support for embeddings enables semantic search, recommendations, and RAG. Vectors are stored alongside structured data, not in a separate system.

SQL support means complex queries — joins, aggregations, window functions — work across all data types.

Tacnode provides integrated search capabilities without a separate search engine.

Full-text search: Inverted indexes, relevance ranking, and text analysis.

Vector search: Similarity search over embeddings with HNSW indexes.

Omni Search: Combine keyword, structured filters, and vector similarity in a single query.

Because search is integrated into the Context Lake, results are always based on current data. No synchronization lag between your database and search engine. No data inconsistency.

Real-Time Analytics

Tacnode functions as a real-time analytics engine with millisecond query latency.

Traditional data warehouses ingest data in batches, achieving good performance but lacking real-time capabilities. Data latency is measured in hours or minutes.

Some systems support real-time ingestion but only guarantee eventual consistency — meaning queries might see incomplete transactions or stale state.

Tacnode delivers:

  • Real-time ingestion: Data becomes queryable the instant it arrives
  • Strong consistency: Full ACID transactions, not eventual consistency
  • Low latency: Millisecond-level interactive queries
  • High concurrency: Multiple workloads operate in parallel

Incremental Materialized Views

Define transformations in SQL. They execute continuously as data arrives.

  • No external orchestration or batch scheduling
  • Features, aggregations, and derived context stay consistent with live state
  • Changes propagate incrementally — no full recomputation

Workload Isolation

Multiple workloads operate against shared context without interfering with each other.

  • Nodegroups: Independent compute resources that scale separately
  • Resource isolation: Analytical queries don’t block transactions
  • Performance guarantees: Ingestion doesn’t slow retrievals

Developer Experience

Tacnode is PostgreSQL-compatible, making it accessible to teams that don’t want to learn a new stack.

  • PostgreSQL wire protocol: Use existing drivers, ORMs, and tools
  • Standard SQL: Your SQL knowledge transfers directly
  • Familiar tooling: Connect with psql, DBeaver, DataGrip, or any PostgreSQL client

Data Synchronization

Built-in data sync from external sources into the Context Lake.

  • Source support: MySQL, PostgreSQL, MongoDB, Kafka, and more
  • Sync modes: Full synchronization, incremental CDC, or both
  • Sharded sources: Consolidate sharded tables into a single table automatically
  • Schema evolution: Handles upstream schema changes — new tables, new columns, type changes