⚛️ QDB (v1.6.0): Quantum-Inspired Deductive Database & Code Hypergraph Engine

QDB (qdb-ai) is a self-contained, multi-hop combinatorial reasoning database, AST codebase dependency hypergraph engine, embedded SQL/OLAP analytics engine, and in-VRAM thermodynamic logit interceptor.

Unlike standard Vector RAG (which relies on shallow 1-hop similarity) or Graph DBs (which suffer from $O(d^k)$ path explosion), QDB formulates multi-hop relational deduction across knowledge graphs and codebases as a Quadratic Unconstrained Binary Optimization (QUBO) problem solved in $< 50 ext{ms}$ via Hamiltonian ground-state energy minimization.

🚀 What's New in v1.6.0

🧠 Adaptive Domain Model Selection: Auto-detects task domain and dynamically loads the optimal HuggingFace model:
- Code: microsoft/codebert-base
- Finance: ProsusAI/finbert
- Science / Research: allenai/scibert_scivocab_uncased
- Medical / Clinical: dmis-lab/biobert-base-cased-v1.2
- Legal: nlpaueb/legal-bert-base-uncased
- General QA: bert-base-uncased
⚡ Automatic Hardware & CUDA Acceleration: Automatically detects NVIDIA GPUs, Apple MPS, or CPU and places model tensors on optimal device.
📦 AST-Level Codebase Ingestion (vault.ingest_codebase): Recursively parses Python AST, JS/TS, package manifests, imports, class inheritance, decorators, and function call chains into k-ary QUBO Hyperedges.
🕸️ k-ary HyperEdges: Multi-participant relational clusters with direct Hamiltonian matrix couplings.
🤫 Clean Output: Suppressed all checkpoint warning clutter.

📦 Installation

pip install --upgrade qdb-ai

⚡ Quickstart

1. Ingest Full Codebase & Query Architecture via QUBO

import qdb
from qdb import Vault

# Initialize Vault
vault = Vault("my_codebase")

# Ingest entire repository (Python AST, JS/TS, requirements.txt, package.json)
report = vault.ingest_codebase("./src/")
print(f"Parsed {report['files_parsed']} files into {report['nodes_created']} AST nodes and {report['edges_created']} hyperedges.")

# Multi-Hop Transitive Dependency Deduction (< 50ms)
res = vault.query("What modules depend on the database layer and what functions are called?")
print(res["answer_narrative"])

2. Multi-Hop Factual Knowledge Deduction

vault.ingest("Alexander Holder received the royal Beryl Coronet as security for a loan of 50000 pounds.", value=50000.0)
vault.ingest("Sir George Burnwell was the secret lover of Mary Holder and masterminded the Beryl Coronet theft.")

result = vault.query("Who masterminded the theft and what was the value of the security?")
print(result["answer_narrative"])

3. In-Memory SQL & OLAP Analytics Over Knowledge/Code

sql_res = vault.query("SELECT id, content FROM states_sql WHERE content LIKE '%FUNCTION%' LIMIT 5")
print(sql_res["sql_results"])

4. In-VRAM Thermodynamic Logit Shield (+50J)

from qdb import Shield
import torch

shield = Shield()
shield.register_ground_truth(
    entity_name="JPMorgan",
    aliases=["JPMC", "$JPM"],
    truth_token_ids=[1042],
    statement="JPMorgan Chase acquired First Republic Bank on May 1, 2023."
)

raw_logits = torch.randn(1, 32000)
penalized_logits, intercepted, telemetry = shield.evaluate("Tell me about $JPM", raw_logits)

📄 License

MIT License. Developed by Prannesshkva.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

qdb_ai-1.7.6.tar.gz (97.2 kB view details)

Uploaded Aug 21, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

qdb_ai-1.7.6-py3-none-any.whl (105.9 kB view details)

Uploaded Aug 21, 2026 Python 3

File details

Details for the file qdb_ai-1.7.6.tar.gz.

File metadata

Download URL: qdb_ai-1.7.6.tar.gz
Upload date: Aug 21, 2026
Size: 97.2 kB
Tags: Source
Uploaded using Trusted Publishing? No
Uploaded via: twine/7.0.0 CPython/3.10.10

File hashes

Hashes for qdb_ai-1.7.6.tar.gz
Algorithm	Hash digest
SHA256	`38934d32599eb7788293875377f3081eb0c92106957ba72bb75636c27705b33a`
MD5	`e57060d6fb1a8afe2982d607e489b6f1`
BLAKE2b-256	`0ac0013ceaa9f261d2338873f45b6f9249b0119ab7687426c14a550ff1b10091`

See more details on using hashes here.

File details

Details for the file qdb_ai-1.7.6-py3-none-any.whl.

File metadata

Download URL: qdb_ai-1.7.6-py3-none-any.whl
Upload date: Aug 21, 2026
Size: 105.9 kB
Tags: Python 3
Uploaded using Trusted Publishing? No
Uploaded via: twine/7.0.0 CPython/3.10.10

File hashes

Hashes for qdb_ai-1.7.6-py3-none-any.whl
Algorithm	Hash digest
SHA256	`fd143d70dd75c44abeb4eef38bdb9fcebef31e893f08f875f283807c30f807ae`
MD5	`d164d263cc8d5e3e93f2d788e892e57c`
BLAKE2b-256	`1360a1881b48876a4a1a6ab28fa9030c446f9d126d7f1ac80e4c7953265a8229`