Our Work

A leading Israeli dance studio was running its entire operation across disconnected spreadsheets — student registrations, class scheduling, billing, and staff management handled manually, with no consolidated view of the business. We designed and built a full operational platform that centralized every function: student lifecycle management from first inquiry through enrollment and payment, class and schedule management across dozens of class types, staff and instructor tracking, automated billing workflows, and real-time analytics dashboards for management.

The result was a dramatic reduction in administrative overhead — staff shifted from manual coordination to student engagement and program development. The system is modular and has scaled with the studio as it has grown.

A leading US-based online auction platform for luxury goods needed executive-level analytics ownership — with no internal data team capable of owning it. We embedded as fractional Head of Data, taking full ownership of the company's data strategy, infrastructure, and business intelligence function.

We maintain and optimize the data warehouse and ETL pipelines, implemented the BI platform used by the entire company, and serve as the primary analytical partner to the executive team. The work spans funnel measurement across acquisition and revenue, marketing attribution, financial forecasting, A/B experimentation, and cohort analysis.

Israel's national dance competition organization was managing thousands of participant registrations, fees, scheduling, and documentation across dozens of studios — entirely by hand. We architected a comprehensive competition management system covering the full lifecycle of each event.

The system manages studio and participant registration, routine submissions, fee calculation and billing (including VAT and invoice generation), scheduling, qualification tracking, and results management. It was built bilingual (Hebrew and English), with integration-ready architecture for SMS and email automation. The system handles competitions with thousands of dancers from dozens of studios across the country.

We built Statsome — a competitive intelligence and digital shelf analytics platform for consumer goods brands. The platform tracked brand presence, pricing, and competitive positioning across digital retail channels, giving CPG marketing and strategy teams a real-time view of their competitive landscape.

The technical architecture — data collection infrastructure, entity cataloguing, and competitive analytics — continues to inform how we approach data product design.

Managing a multi-strategy TASE portfolio manually — tracking dozens of symbols, timing buy/sell decisions across independent strategies, and enforcing risk rules consistently — doesn't scale. We built a fully autonomous trading platform that runs multiple AI-driven agents in parallel, each operating its own portfolio according to a distinct strategy and risk profile.

Each agent is powered by a Claude language model that runs on a configurable cycle: fetching live market data, evaluating every watched symbol against current positions and portfolio health, and issuing BUY/SELL/HOLD decisions with a confidence score. A rules-based risk manager sits between the AI and execution — enforcing position size limits, stop-losses, daily loss halts, and drawdown thresholds before any order goes through. The platform currently operates in virtual mode for strategy testing and backtesting, with architecture in place for live execution via Interactive Brokers.