Things we've built, end to end.
A studio's portfolio is its argument. These are the systems that came out of Guru Meditation — enterprise SaaS, whitelabel broker infrastructure, autonomous trading, and applied AI we run on our own hardware. What ties them together isn't a stack — it's that each one solves a real business problem we understood before we wrote a line of code, then built from the database schema to the screen. Open any project for the full story.
Response365
An entire business stack in one place — roughly 80 mounted apps behind 16 role-based dashboards and around 1,900 screens, spanning finance, sales, supply chain, manufacturing, HR and compliance. One multi-tenant database, AI woven through the workflow.
read the case study →Business Intelligence
A conversational analytics layer living inside Response365: ask in plain language, in any language, and a ten-stage validated query pipeline answers across every module — real-time role KPIs, auto-chosen charts, Excel/CSV/PDF export. Self-learning, not a thin LLM wrapper.
read the case study →Food Production & Security
A food-native manufacturing execution system — versioned recipes, electronic batch records, batch genealogy that posts straight to the ledger — paired with a two-layer food-safety platform: HACCP, allergen control, one-click multi-jurisdiction recall and supply-risk scoring.
read the case study →Piggy Broker
A brokerage in a box. FIX 4.4 liquidity aggregation across multiple LPs, multi-LP best-bid/offer, best-execution routing, A-book / B-book / mixed handling, an internal price-time matching engine, a dealing desk, live net-exposure, auto-hedging and a margin engine with stop-out.
read the case study →Piggy Trader
Not a rule-based bot — a meta-controller that learns which kind of market favours which kind of strategy across price, volume, volatility regimes, intermarket and macro signals, and sizes every position by volatility and edge. Strategies ship only after walk-forward, out-of-sample validation.
read the case study →Piggy Native & Mobile
Native desktop and mobile trading terminals wired straight into the broker backend — the same order book, quotes and execution flow on the desk and in your pocket. A Rust-core desktop app with signed installers, and a React Native mobile app with biometric login and push.
read the case study →Universal Content Extractor
Pulls stories from PDF newspapers and magazines, websites, RSS/Atom and social platforms, then segments, embeds and semantically clusters coverage across languages and ranks it into a personalised feed. Every LLM step runs on a self-hosted Gemma model locally — the angle is cost.
read the case study →Trading Sentiment
Reads articles coming out of the content-extraction engine, resolves the companies and tickers behind each one, scores it bullish or bearish with a local LLM, and turns it into live BUY/SELL/HOLD calls for stocks and LONG/SHORT for forex — streamed over WebSocket and SSE sub-second.
read the case study →CreditHelper
An end-to-end loan lifecycle platform for lenders — origination, flexible terms, automated schedules, a regular-and-late interest engine, batch invoicing and multi-channel reminders. GoCardless direct debit collects and reconciles automatically, matching payments even without a reference.
read the case study →Live Dating Platform
A members-only dating product on one backend with two clients — a Next.js web app and an Expo/React Native app over a Bearer API. In-app WebRTC video & voice calling, AI image moderation, distance-and-preference matching, realtime chat and Stripe tiers, wrapped in a trust-&-safety stack — private media, magic-byte validation, ban/block and KYC-ready verification — built in from the schema up.
read the case study →