Flagship R&D / Code name only
Project AvaTrader — Code Name
Project AvaTrader (code name) is Codepact’s active quantitative trading technology R&D initiative. ‘AvaTrader’ is an internal project code name—not a final product or company identity.
A controlled technology system for turning market hypotheses into auditable decisions.
The project explores an end-to-end foundation for quantitative trading research, validation, risk, execution, evidence, and future AI-enabled decision support.
It is more ambitious than a simple trading bot. The goal is a modular system in which every proposal passes through explicit evidence and safety boundaries before controlled execution is even considered.
The programme is in active research and engineering. This page does not claim live-trading readiness, guaranteed profitability, or completed future capability.
Public system concept
Intelligence earns authority one boundary at a time.
This is a conceptual communication of system responsibilities. It intentionally excludes strategy rules, private architecture, broker details, credentials, and proprietary datasets.
Research should be able to reject a weak hypothesis. Risk should be able to constrain a promising one. Execution should not be able to bypass either. Every stage should create evidence for audit and learning.
AI direction sits inside this structure. Adaptive intelligence may assist research and decision-making, but deterministic safety and execution constraints remain explicit.
Problem / opportunity
Financial markets expose weak engineering quickly.
Research can overfit
A convincing backtest can reflect selection, leakage, or fragile assumptions rather than a durable market hypothesis.
Risk can become implicit
If safeguards live inside strategy logic, they become difficult to reason about, test independently, or enforce consistently.
Execution changes the result
Latency, costs, market conditions, and failure behaviour can separate theoretical opportunity from operational reality.
AI can exceed its evidence
Autonomy without boundaries can amplify a weak premise faster than it can create a reliable decision system.
Research & safety
Reject, constrain, validate, then advance.
Demo validation comes before any separately approved live deployment.
Research promotion and deployment approval are different decisions. A strategy or system component can progress through research without receiving authority to operate live.
Future live use, if pursued, would require separate validation, operating controls, and approval. Nothing on this site represents such approval.
Development journey
What exists today and what the next stage is designed to deepen.
Active research and engineering around a coherent system foundation.
The present programme establishes the research discipline, architectural separation, risk philosophy, and evidence model needed to evaluate the larger commercial thesis.
Expanded quantitative research
Deeper platform engineering
AI integration inside defined boundaries
Multi-market development
Infrastructure and validation
Commercialisation planning
Investment pathway
Fund the next stage, not a manufactured promise.
Codepact is seeking an initial US$1.5 million growth and R&D investment. This is the initial raise for the next stage—not a statement of total lifetime capital requirements.
Review the investment pathway