Foundation · 02 of 14

ONI is the reason the other nine agents are never wrong.

The foundational data lakehouse + internal logic engine of JOY Automart. ONI does not make business decisions. Its only job: ingest unstructured chaos (clicks, supplier PDFs, fitment logs) and weave it into a flawless, queryable narrative. JOYADH asks. ONI answers. The army acts.

Data Infrastructure & Lakehouse·The bedrock · absolute contextual memory
Mission

Maintain the four JOY Automart narratives — Vehicle & Fitment · Customer Behaviour · Supply Truth · Margin Reality. Provide every other agent a single, instantly-queryable Truth™ on demand.

What it does today

Capabilities · live in production.

  • 01

    Continuous data pipelines

    Clean dirty supplier APIs · categorise customer behaviour · normalise scrapes into pricing graphs.

  • 02

    Semantic layer

    When JOYADH needs a decision, ONI returns a unified data packet linking past purchases, current warehouse stock and weather context.

  • 03

    JOY VIN fitment matrix

    Hundreds of thousands of parts mapped to exact Make · Model · Year · Engine Trim.

  • 04

    Behavioural profiles

    Tracks every hover, abandoned cart, search query. Builds the unified profile that ZIFA personalises against.

  • 05

    True landed cost

    Logs every JASMINE movement. Calculates the real cost of a brake pad factoring in hidden shipping delays.

  • 06

    Bilingual entity resolution

    Bengali ↔ English semantic match so a slang query 'চিমটা' returns 'Lower Control Arm'.

Power

Absolute contextual memory

ONI is the reason the system learns. Every ZIFA ad result, every ARES combat brief, every JASMINE fulfilment outcome — ONI records it. The next JOYADH decision is smarter because of it. The architecture compounds.

Weakness · by design

Action paralysis · by design

ONI is strictly observational. Even if the data clearly shows a mispriced transmission fluid, ONI cannot change the price. It must wait for JOYADH to read the data and order ARES or ZIFA to act.

Architecture

The stack that powers it.

LayerTechPurpose
Relational anchorPostgreSQL + pgvector / MariaDB VectorACID-compliant SKU counts · VIN fitment · supplier POs · ledgers
Vector brainpgvector · Pinecone / QdrantSemantic similarity for behavioural profiles · RAG memory
LakehouseDatabricks · Snowflake · S3 + IcebergCheap dump for raw HTML scrapes · supplier PDFs
Hybrid queriesSingle SQL · vector + relationalFind semantically relevant parts WHERE inventory>0 AND margin>15%
Synergy

Who it talks to · what it sends.

  • → JOYADH
    respond
    Combat narrative · behavioural packet · live inventory
  • → ARES
    feed
    Cleaned competitor pricing graphs · historical drop patterns
  • → ZIFA
    feed
    Customer segment profiles · purchase history · churn signals
  • → JASMINE
    log
    Every supply-chain movement · landed-cost reconciliation
  • → AAJ
    feed
    Ledger-grade transactional truth for books closing
Next builds

How we're extending it.

  • Realtime CDC pipeline
    Change-data-capture so every Mongo write streams to the ONI lakehouse within seconds.
  • Embedding refresh job
    Nightly re-embedding of new SKUs + behavioural events so semantic search stays current.
  • Predictive layer
    Forecast competitor sales based on historical drop cadence · feed JOYADH 48h before they trigger.
  • Open lake export
    Investor / auditor read-only export of cleaned narratives via Iceberg tables.
Try it

Want to see ONI in action?

© 2026 JOY Automart · Dhaka, BD · A 14-agent Sovereign OS on Claude Opus 4.8.