About/Stealth
A research lab for the
machinery of modern markets.
NeuralTrading.group is a small, technically-led team building the models, execution engines and infrastructure that institutional capital runs on. We work across DeFi arbitrage, systematic trading, AI agent systems and applied research — and we are in stealth while we build.
Mission/Why we exist
Markets are won with better research and better engineering. We build both — rigorous quantitative work, shipped as production-grade systems — so that serious capital can act on it with confidence.
We are not a fund and we are not a vendor of black boxes. We are a research and engineering lab: we study how markets behave, turn that into models and tooling, and build it to a standard that holds up to institutional scrutiny.
Approach/How we work
A few principles shape how we engage and what we ship.
Research first
Every product starts as a question, not a pitch. We study signals, market structure and on-chain behaviour before we write the code that trades on them.
Embedded, senior partners
We engage as a small, senior team beside yours — scoping the problem, building the work, and staying close while it is validated before it touches capital.
Institutional-grade engineering
Real money demands real software. We work with tested, observable, access-controlled systems and treat security, auditability and reliability as requirements.
Open about evidence
Where we can, we publish and reproduce our work. We would rather show how something holds up than ask you to take our word for it.
Process/How an engagement works
A working relationship you can see into.
Every engagement runs through the same five phases. Each one ends in something concrete you can hold, review and take with you — so the work stays legible from the first conversation to the last, and you are never asked to take progress on trust.
Scope
We sit with your team to define the problem, the constraints and what “good” would look like — before any code is written.
You get: a written scope and engagement brief — the problem framed, assumptions made explicit, success criteria agreed.
Research
We study the signals, market structure and on-chain behaviour the problem depends on, and test whether the approach holds up before committing to a build.
You get: a research note — the methods we propose, why they fit, the trade-offs we see, and an honest read on what is feasible.
Build
We engineer the models, execution logic and infrastructure as tested, observable, access-controlled systems — built to survive institutional scrutiny.
You get: working software in a repository you can read, designed for auditability — not handed over as a black box.
Validate
We exercise the work the way it will actually be used — backtests with cost modelling, testnet and integration runs, static analysis and security hardening — before anything touches live capital.
You get: a validation record — what was tested, how it behaved and where the limits are, reproducible rather than asserted.
Hand-over
We transfer the work into your environment with the documentation and context your team needs to own it — and we stay close while it settles in.
You get: the codebase, documentation and operating context to run it yourself, with knowledge transfer to your engineers.
Discreet by default · senior throughout · nothing reaches capital before it is validated
Capabilities/What we do
Four areas, one discipline.
We concentrate where quantitative research and serious software meet. The verifiable detail — real repositories, with honest build stages — lives on the products page.
01—
DeFi Arbitrage
Multi-chain, MEV-aware detection, optimization and on-chain execution.
02—
Quant & Algo Trading
Systematic research and automation, from signal to execution and risk.
03—
AI Agent Infrastructure
Orchestration, governance and durability for multi-agent systems.
04—
Applied Research
Reproducible, evidence-led work, published as a companion.
Who we serve/Audience
Built for institutions that take engineering seriously.
We work with a small number of institutional partners at a time. The teams we are built for include:
Banks
Markets, treasury and innovation teams modernising trading and analytics.
Asset & wealth managers
Systematic strategies, portfolio tooling and explainable decisioning.
Hedge funds
Alpha research, execution engineering and risk infrastructure.
Prop-trading firms
Low-latency execution, automation and on-chain strategy work.
Fintechs
Trading, wealth and DeFi capabilities built into your platform.
Family offices & HNWI
Discreet, research-driven wealth technology and bespoke solutions.
Backed by/AIgen Solutions
An AIgen Solutions company.
NeuralTrading.group is the quant and DeFi research practice of AIgen Solutions S.r.l. — the engineering company behind our models, platforms and infrastructure. We bring AIgen's “humans and algorithms” approach to institutional markets.
Visit AIgen SolutionsWorking on something serious? Let's talk.
We are taking on a limited number of institutional partners while in stealth. If you are building in DeFi, systematic trading or wealth tech and want a research-driven team beside you, get in touch.