Polysnipes.com - Product, Data, AI, and Trading in One Terminal

Turning prediction-market noise into a research workflow

PolySnipes gives Polymarket researchers one workspace for outcomes priced at 5¢ or below. The live core covers discovery, analysis, wallet access, and signed execution. Moonshot Legends, now in final testing, adds public-trader research. Its 24-hour-delayed preview is live.

Researchers can find overlooked markets, compare the quote with their own probability estimate, model risk and exit scenarios, review ranked candidates, and sign a trade.

I owned the product from positioning and interaction design through frontend, backend, market-data infrastructure, AI, wallet authentication, trading, and Web3 payments.

I used one rule throughout the build: show the evidence before asking the user to act. A low price alone does not support a trade.

Polymarket has a large market universe that changes throughout the day. Researchers studying low-price outcomes must move between market pages, calculators, wallet tools, and activity feeds to evaluate a single quote.

PolySnipes had to make thousands of outcomes fast to search and compare. It also had to connect research with wallet identity, execution, and payment controls without presenting a forecast as a fact.

The interface, data pipeline, AI layer, wallet state, and transaction flow needed to work as one system across desktop and mobile.

I built a shared terminal shell so each view carries market context into the next task.

Scanner indexes outcomes and supports search and sorting by price, volume, liquidity, and expiry.

Analyzer turns a selected quote into expected value, trade-exit scenarios, and Kelly-based position sizing.

Portfolio connects a PolySnipes account to the user’s Polymarket wallet for balances, positions, and signed execution.

3D Agent ranks five markets for review and maps rank to physical depth.

The live scanner indexes thousands of outcomes in one session. List virtualization keeps search and scrolling responsive across the dataset.

I restricted the Agent to two tasks: rank a supplied market set and write a short research hypothesis. The model cannot provide prices, liquidity, expiry dates, URLs, news, polling, or probability estimates.

The backend validates the response against a schema, rejects unknown market IDs, and rebuilds each displayed fact from PolySnipes records. A deterministic ranking takes over when a provider fails or returns invalid output.

These controls keep the Agent useful without treating model output as market data.

I designed the Web3 flow around user approval at each step.

The backend derives the user’s Polymarket Safe, exchanges signed wallet headers for CLOB credentials, reads the order book, estimates the fill, validates the Polygon chain, and prepares the order. The user reviews and signs each live action.

PolySnipes does not run automated strategies or hold user funds. The interface explains those boundaries beside the relevant controls.

Moonshot Legends studies public traders whose campaigns started at 5¢ or below. I released the feature teaser while the full ledger entered final testing.

I built the scoring and accounting pipeline from public wallet activity:

FIFO campaign accounting

Wilson lower-bound hit-rate scoring

ROI and profit-factor normalization

repeatability and activity-span checks

realized P&L and maximum-drawdown controls

sample-confidence penalties and exclusion reasons

The interface displays the methodology version, sample size, freshness, and gross/pre-fee status. The public preview delays data by 24 hours and withholds wallets, histories, live trades, and alerts.

At launch, subscribers receive the Top 100 ledger, trader dossiers, a public activity feed, up to 20 follows, and Telegram alerts. I also built a feature-gated entitlement flow for 30-day and 90-day passes. The backend approves one pUSD transfer payload, verifies its Polygon receipt and confirmation count, then grants timed access. The passes do not renew on their own.

A screenshot captures one win. Moonshot Legends accounts for the trader’s full qualifying record, losses included. The teaser presents that idea as a separate motion project on my Contra profile. This case study covers the system behind it.

I based the interface on statistical bureaus, market ledgers, and printed research instruments. The design avoids the neon palette and decorative glass common in Web3 products.

The system uses paper, graphite, signal orange, one-pixel rules, tabular type, square controls, and dense layouts. Color marks actions and state, while labels carry the same information in text. Loading, stale, empty, entitlement, and error states sit inside the workspace.

The visual hierarchy keeps market evidence, system state, and the next available action in view.

Responsive research terminal in production

Shared market-data pipeline for Scanner and Agent

Virtualized scanner for thousands of low-price outcomes

EV, trade-exit, and Kelly analysis modes

AI ranking with schema validation and trusted-data reconstruction

Privy authentication and Polygon wallet integration

Polymarket Safe, CLOB, order-book, portfolio, and signed execution flows

24-hour-delayed Moonshot Legends preview

Public-trader scoring and campaign accounting

Telegram follow and alert infrastructure

pUSD payment verification and timed entitlements

Top 100 ranking, dossiers, live activity, and paid access

I shipped the PolySnipes core as a live Web2 + Web3 product. It connects market discovery, quantitative analysis, constrained AI, wallet authentication, and signed execution in one interface. Moonshot Legends extends that system with public-trader research and a paid access layer.

The work reflects my role as a Full-Stack Product Builder. I set the product boundary, designed the interaction model, built the interface and backend, integrated external protocols, and added the controls required for live financial workflows.

Live product: https://www.polysnipes.com/