Case study · Backtesting platform for investment strategies

A system for testing and validating investment strategies on the markets

In short: a system for testing and validating investment strategies across crypto, FX and indices. Primessio designed, built, deployed and maintains it on FastAPI, PostgreSQL, Next.js and Docker. Only a strategy with a confirmed edge goes live — after a history backtest, live measurement and an out‑of‑sample robustness check, under automated 24/7 risk control.

Tests strategies on the markets (crypto, FX, indices): signals from the real market → backtest and paper‑trading → evidence validation → automated execution under risk control. Only a strategy with a confirmed edge goes live. Designed, built, deployed and maintained by Primessio — one of many applications we have shipped.

Role
Design · Build · Deploy · Maintain
Type
Backtest + paper‑trading + validation + auto‑execution
Stack
FastAPI · PostgreSQL · Next.js · Docker
Markets
Crypto · FX · indices

Product demo

Anonymised screens — the same templates and styles as production, sample data. Autoplay · PL / DE / EN. · Open the demo in a new window ↗

The challenge

Testing investment strategies usually ends at a spreadsheet and a single chart. A genuine edge is hard to tell from luck, drift is easy to miss, and hand maths can't keep up with a market that trades 24/7.

The client needed a system that tests strategies on real market data, lets ONLY the ones with a proven edge go live, and acts automatically — guarding risk itself instead of relying on the operator's intuition.

The solution

Primessio designed and built one platform covering a strategy's full lifecycle: signal intake from the real market (crypto, FX, indices), backtest on history and paper‑trading live, a multi‑stage validation gate (only a strategy with a confirmed edge goes live), automated execution, and built‑in risk control with a kill‑switch.

Plus full observability (dashboards, result curves, 24/7 readiness reports), continuous self‑improvement („challenger” variants tested on data) and an audit of every decision. All containerised, with data in the EU.

Scope — what we delivered

Real‑time analytics

Real‑time data and signal intake, normalisation and live dashboards.

Decision engine

Every decision through a multi‑stage evidence gate — no proof, no action.

Evidence validation

Backtest + live + robustness (out‑of‑sample) — luck is ruled out.

Automated execution

Acting 24/7 with no human, idempotent, with full feedback and audit.

Risk control

Size and exposure limits, daily stop, portfolio control, kill‑switch.

Position management

Automatic protection of hard‑won results (trailing stop) — parametric and tested.

Self‑improvement

„Challenger” variants, A/B tests and periodic reviews — the system learns from evidence.

Observability & audit

Role dashboards, 24/7 readiness reports, a log of every decision and operation.

Outcomes

These are operating targets and system characteristics, not a profit promise — every figure follows from evidence discipline (backtest + live + robustness), not a forecast.

+10% monthly target — a result curve under risk control
3 proofs action only with confirmation: history + live + robustness
24/7 automated action and risk control — no spreadsheets
0 „by feel” decisions — each has proof and an audit

Why Primessio

A team with 20+ years in IT — 50+ rollouts, incl. for large brands and public institutions

Full ownership: we design, build, deploy and maintain (24/7)

Evidence discipline — only a confirmed‑edge strategy goes live

Data in the EU, no vendor lock‑in

How we work — agile

Go‑live is the beginning, not the end. We work in short cycles, continuously adapting the system to real data and optimising the result — with evidence discipline at every step.

01

Discovery

We map decisions, data and risk rules — before the first model.

02

Design

Data architecture, decision gates and UI — for real time and cost.

03

Build

Short sprints, fast feedback, empirical verification of every change.

04

Go‑live

Production, live data, automation under risk control.

05

Run

Variants, A/B and reviews — the system keeps raising the result.

Technology

Deliberate choices for reliability and cost: containerised services, data in the EU, background automation (24/7 scheduler), no vendor lock‑in. Every decision is backed by evidence — the goal does not bend the data.

  • Backtesting
  • Quant platform
  • FastAPI
  • PostgreSQL
  • Next.js
  • Docker
  • 24/7 scheduler
  • OOS validation
  • Risk control
  • Audit
  • PL · DE · EN

Frequently asked questions

What is a system for testing investment strategies?

A platform that tests strategies on real market data (crypto, FX, indices): it generates signals, runs a backtest on history and paper‑trading live, and lets a strategy go live only after a multi‑stage evidence gate (history backtest + live measurement + out‑of‑sample robustness). Execution is automatic, under built‑in risk control and with a full audit.

What do „evidence‑based decisions” mean?

No candidate goes into action on real resources until it has proven its edge independently: on a full‑history backtest, in live measurement and in a robustness check (out‑of‑sample, anti‑noise). This eliminates false positives and „by feel” decisions.

How does the system guard risk?

Risk is built in: size and exposure limits, a daily stop, portfolio correlation control, automatic protection of hard‑won results (trailing stop) and a kill‑switch. It runs 24/7, independent of human attention.

Do we need expensive infrastructure?

No. The system is containerised and runs on an ordinary server: minimal dependencies, no vendor lock‑in, data in the EU. Low running costs were one of the project requirements.

What happens after go‑live?

Go‑live is the start of the partnership. We run and evolve the system in an agile model: „challenger” variants, A/B tests and periodic reviews keep raising the result — and every change is verified empirically.

Who designed and built this system?

Primessio (a brand of ACTIVE IT) designed, built, deployed and maintains it — a team with 20+ years in IT projects. It is one of many applications we have shipped.

What stack does it run on and which markets does it cover?

The stack is FastAPI, PostgreSQL, Next.js and Docker; services are containerised, data in the EU. It covers crypto, FX and indices, running 24/7 under automated risk control.

Are the strategies backtested?

Yes. Every strategy goes through a full‑history backtest, live measurement (paper‑trading) and an out‑of‑sample robustness check; only a strategy with a confirmed edge goes live.

Got strategies you want to test and only take live with proof?

We will design, build, deploy and maintain a system that tests your investment strategies on real market data and lets only the ones with a confirmed edge go live — with risk control and 20 years of team experience.

Let's talk