Harry Dhindsa.

QuantDesk — AI Quant Team

Building AI Teams for Portfolio Managers

Designing and deploying autonomous AI research teams that replicate the structure of an institutional quant desk — operating 24/7 across the full research and trading lifecycle. Have your own team of AI agents covering the entire spectrum of systematic research, development, and execution.

• Quant Researcher: Mathematician and signal architect constructing structured trading strategies grounded in statistical theory, financial research, and economic intuition. Develops hypotheses, formalizes signal definitions, and designs robust testing frameworks.
• Quant Developer: Implements strategy specifications into production-grade backtesting frameworks. Ensures correct signal lagging, cost modelling, walk-forward validation, and reproducible research pipelines.
• Quant Trader: Oversees live and paper strategy monitoring. Evaluates real-time performance versus expected distributions, tracks execution quality, and assesses regime shifts or performance drift.
• Data Analyst: Maintains data integrity across asset classes. Sources, cleans, normalizes, and structures market and research datasets for downstream modelling and signal construction.
• Risk & Validation Officer: Independent audit layer enforcing statistical discipline. Reviews strategy robustness, detects overfitting risk, validates out-of-sample integrity, and stress-tests assumptions.
• Orchestrator: Coordinates the full AI research pipeline. Manages agent workflows, scheduling, logging, and reporting — ensuring continuous research velocity and structured output.

Quantitative Trader / Researcher

• Implemented bespoke systematic systems and dashboards for Portfolio Managers running Macro Discretionary Strategies at Pod shops.
• Built probability-based algo for Polymarket BTC binary markets using diffusion models, probability calibration, and order-flow microstructure. View Dashboard →
• Running Macro RV, Cross-Sectional Momentum, Stat-arb Systematic Strategies. PCA, HMM, MS-VAR, CPD, LSTM. See Strategies →
• Single stock quantamental strategies.

Prediction Markets Analytics Platform pmanalytics.vercel.app →

• Implemented probability distribution modelling, order-flow imbalance metrics, cross-platform arbitrage scanners, and anomaly detection across binary event markets.

HDAlgos hdalgos.com →

• Built and shipped 25+ proprietary TradingView indicators, integrating trend models, regime detection, PCA-based residual signals, and volatility systems.
• Built the full web platform end-to-end (React, Tailwind, Supabase, Stripe, Vercel) including authentication, subscription infrastructure, and API integrations.
• Released an open-source indicator which became one of TradingView’s fastest-growing scripts within 24 hours.

Zayalana zayalana.com →

• Implemented LLM chatbot (Claude Sonnet 4.5), including embeddings, memory modules, and context retrieval via RAG architecture.
• Python-based scraping engines to extract data across multiple marketplaces (Shopify, TikTok, Amazon, AliExpress).

WhoWereYou App Store →

• Find Mode: Post anonymous descriptions of someone you noticed in real life — location, time, details. The other person can claim the sighting, and if both confirm, you match.
• Play Mode: Browse real profiles and curate a shortlist of up to five people you’re genuinely interested in.
• Built-in safety features: anonymous by default, content filtering, reporting tools, and in-app safety guidance.