Explore
Follow interesting ideas and technical questions before deciding whether they are worth pursuing.
Exploring ideas before they become technologies.
QuantsMind Labs is the exploration space for experiments, prototypes, technical investigations and emerging ideas across the QuantsMind technology ecosystem. Labs explores new ideas, prototypes technologies and validates technical directions before they are committed to a product or commercial offering.
Labs gives QuantsMind a place to explore technical possibilities before they harden into products. It is a working philosophy — a way of directing curiosity, not a promise about any single project.
Follow interesting ideas and technical questions before deciding whether they are worth pursuing.
Build small, honest prototypes to test whether an idea works in practice — not just on paper.
Check assumptions against reality, keeping scope clear and claims measured.
Capture what works and what does not — even abandoned directions teach the next one.
Let mature experiments find a home: a technology, a product, or the wider engineering work of QuantsMind.
An idea can move through these stages — or stop at any point. Not every experiment reaches the final stages.
Open research code published under the QuantsMind GitHub organisation. Every project below links to its own public repository, so what is claimed here can be checked against the source. None of these is a finished product, and none should be relied on for production work.
A vector database written in Python, exploring approximate nearest-neighbour search, write-ahead logging, sharding and replication. Roughly 7,400 lines across 66 modules with a substantial test suite. Licensed MIT.
Not established. Single-process only. There is no cross-node transport, consensus, leader election, failover or shard migration, so the distributed architecture described in its README is a roadmap rather than a capability. The published benchmark figures were measured on a 2,000-vector, 32-dimensional toy dataset. Tagged versions run ahead of the version declared in the package manifest, and no release has been published.
An open-source-intelligence collection and analysis platform with a Python backend, a Vue frontend, role-based access control and an audit middleware that redacts sensitive payload keys. Ships container and Kubernetes manifests.
Not established. Experimental. An exploratory initiative investigating open-source intelligence collection, analysis and information workflows. The project remains under feasibility evaluation and is not offered as a production-ready product. The backend test suite has not passed, and there is no licence file or release.
A feasibility study into classifying news text, containing a cleaning pipeline, a training script, an evaluation harness and a small HTTP prediction service, with a Dockerfile and Python packaging configuration.
Not established. Its reported accuracy is not independently validated and must not be relied on. The committed evaluation report shows a perfect score, but that score was measured on a synthetic dataset generated from a handful of title and body templates, which any classifier separates perfectly; it does not measure real-world fake-news detection. No trained model checkpoint is included. When none is present the service loads an untuned base model with an untrained classification head, so its current prediction behaviour is not reliable evidence of model performance. There is no licence file and no release.
These are research projects and prototypes, not completed products. A project may change direction, remain experimental, be incorporated into another technology, or be discontinued.
Labs provides a place to explore technical possibilities before committing them to a product or commercial offering.
The pipeline is directional. Ideas enter at the idea stage and move forward as they gain confidence — or are stopped while still early.
Some exploration has already matured into concrete, public technologies. These represent where QuantsMind exploration has produced tangible results.
A concrete QuantsMind technology: an open Python SDK for building, executing and analyzing quantum programs, published on PyPI at version 1.1.0. Execution is by simulation; no quantum hardware backend is included.
A concrete QuantsMind technology: a statically typed systems programming language and computing ecosystem with a self-hosted compiler, published at version 1.1.0. The language and compiler are early and the API surface may still change.
An open, vendor-independent scientific computing SDK, published at version 1.1.0, with functionality across mathematics, the sciences, finance, AI agents and knowledge graphs. It is a young project with a rapidly growing surface; coverage and numerical accuracy vary by domain.
Beyond today's projects, Labs keeps an eye on broader directions. These are areas of interest — not commitments and not completed products.
QuantsMind does not have products in every category listed above. These are the areas the Labs exploration surface keeps open.
Exploration can feed the wider QuantsMind ecosystem. The diagram below is conceptual — it shows a direction, not a claim that every project follows this path.
Not every Labs project becomes a product. This flow shows how experimentation can feed the wider QuantsMind ecosystem over time.
Labs work is exploratory. Projects may change direction, remain experimental, be incorporated into other technologies, or be discontinued.
From SDKs and languages to the ideas still forming, the QuantsMind ecosystem is open, experimental and underway.