Xenos – AI-Assisted Trading Research
Xenos is building a Google Cloud-powered AI-assisted backtesting and trading research platform for independent traders. The platform focuses on BTC and XAU historical strategy research with honest, verifiable results.
Xenos is currently a working MVP under active development by founder Ripon Mallick, based in India. Contact: riponmallick@xenos.sbs
Product — Working MVP
The current Xenos MVP processes historical BTC and XAU market data, calculates technical and market-structure features, runs strategy backtests and produces performance reports including returns, drawdown, win rate and profit factor. A web dashboard for user-configured experiments is currently under development. Live trade execution is not part of the MVP — the platform is for research and paper-trading use only.
- Import Historical Data: Load BTC or XAU historical datasets into the research pipeline. Exchange-account connections are not part of the current MVP.
- Configure an Experiment: Select strategy parameters, risk settings and the historical period to evaluate. A visual strategy builder is planned for the private beta.
- Run a Backtest: Generate simulated trades, an equity curve, win rate, profit factor and maximum drawdown using historical data.
Structured historical backtesting and machine-learning research tools for independent traders.
Google Cloud Architecture
Xenos uses Google Cloud to separate the web application, historical data storage, analytics and compute-intensive workloads.
- Cloud Storage — Experimental. Currently being configured to store historical BTC and XAU datasets and generated research artifacts.
- Cloud Run — In Development. Planned to host the Xenos web API and containerised backtesting workers.
- BigQuery — In Development. Planned to analyse completed backtest trades, strategy performance and model predictions.
- Compute Engine — Planned. Planned for larger batch backtests and model-evaluation workloads after Cloud Run MVP is stable.
- Vertex AI — Planned. For managed ML experimentation, experiment tracking and model evaluation after the core private-beta workflow is stable.
- Firestore — Planned. For user profiles, saved strategies, backtest job status and subscription data.
How Google Cloud Support Will Be Used
Resources will be used to deploy the API and backtesting services, store and process BTC and XAU datasets, execute research jobs, analyse trades in BigQuery, support private-beta users, evaluate ML forecasting approaches, and implement workload monitoring and cost controls.
Working MVP
Status: Working MVP under active development. Cloud deployment work is in progress: Cloud Storage is experimental, and Cloud Run and BigQuery integrations are in development. Private beta is being prepared for 25–50 independent traders with a target of 500+ completed backtests.
Published Reproducible MVP Evidence
Xenos published a BTCUSDT M15 simulated historical baseline covering 1 January through 30 June 2026. The long-only EMA 20/50 experiment processed 17,376 public candles with no detected interval gaps and produced 164 completed trades.
- Dataset integrity: 17,376 candles, zero detected interval gaps and a published SHA-256 hash.
- Execution evidence: 164 completed simulated trades with next-candle-open execution.
- Cost modelling: Fees and slippage are included on both entry and exit.
- Public artifacts: Source candles, configuration, completed trades, summary, equity curve and reproduction script.
The full evidence package discloses all performance metrics and trades, including unsuccessful results, rather than presenting a profitability claim.
Simulated historical research result. No live orders were placed. It does not represent live trading performance and is not financial advice.
Six-Month Roadmap
- Phase 1 — Stabilize the Cloud MVP: Stabilize the existing backtesting workflow, deploy the API on Cloud Run, connect Cloud Storage and BigQuery, improve dashboard reporting, add monitoring.
- Phase 2 — Private Beta: Add authenticated accounts, asynchronous jobs, recruit 25–50 beta testers, complete 500+ backtests, conduct 10+ customer interviews, validate first paid subscription.
- Phase 3 — ML Research: Introduce Vertex AI, compare at least two ML forecasting approaches, test across market regimes, add experiment tracking, improve reliability and cost controls.
About
Founder: Ripon Mallick | Founded: January 2026 | Stage: Working MVP under active development | Market: Independent trading research | Infrastructure: Google Cloud | Location: Chandannagar, West Bengal, India | Contact: riponmallick@xenos.sbs
Apply for Private Beta
Business Model
Xenos plans to operate as a subscription software platform. The initial offering will include a limited free research tier and a paid individual plan with additional backtesting capacity, saved experiments and advanced model evaluation. A future team plan may support small research groups and API access. Pricing will be validated during the private beta.
Funding Plan
Xenos is currently bootstrapped by its founder. After validating the product through private-beta usage and customer interviews, Xenos plans to explore pre-seed funding and strategic startup partnerships to support product development, infrastructure and customer acquisition.
Financial Disclaimer
Xenos is a research and paper-trading platform. It is not a brokerage, investment adviser or asset-management service. All backtest results are simulated historical research results. They do not represent live trading performance and are not financial advice. Trading involves significant risk. Past performance is not indicative of future results.