Operational Systems

Systems

A curated index of real systems built from first principles. Each one started as a problem worth solving.

System 01

Foretyx

Privacy Infrastructure · Security Layer

Problem

Indian digital users lack a unified, privacy-first infrastructure layer. Data leaks, opaque permissions, and zero user control are the norm across apps and services.

Approach

Building Foretyx as a platform-level privacy layer — not just a tool. A security substrate that any Indian application can integrate to give users real control over their data perimeter.

Vision

Privacy infrastructure as a platform play. Foretyx becomes the security layer for every Indian application that takes user trust seriously.

Privacy Tech Security India-first Platform
smartphone
desktop_windows
shield
cloud
storage
Foretyx — Privacy Layer Architecture

schemaArchitecture

Foretyx sits as a middleware security layer between applications and data stores. It intercepts data flows, enforces user-defined permissions, and provides an audit trail — all without modifying the host application's core logic.

flagStatus

Currently in early design and feasibility phase. Core privacy primitives being defined. Target market: Indian SaaS companies and consumer apps handling sensitive user data. Open to collaborators.

smart_toy
smart_toy

Local LLM · Browser-Native · Ollama

System 02

AI Browser Chatbot

Local LLM · Ollama · Browser Access

Problem

Cloud-based AI assistants send your data to remote servers. Most people cannot self-host LLMs. There's no simple, browser-based AI that runs fully local with real browsing capability.

Approach

Browser-based chatbot powered by Ollama running local models (DeepSeek, LLaMA). The UI lives in the browser, the model runs on-device. No API keys, no data leaving the machine.

Outcome

Fully private, zero-cost AI assistant with real browser context awareness. Works offline. No subscriptions.

Ollama DeepSeek Local LLM Browser

memoryArchitecture

Frontend HTML/JS chatbot interface communicates with Ollama's local REST API. The model runs as a background process. Browser extension layer adds page context — the model can read and reason about the current page.

balanceTradeoffs

Requires Ollama installed locally — adds setup friction. Response speed depends on hardware. In exchange: zero cost, total privacy, full offline capability. Worth it.

System 03

Smart Home Automation System

Intelligent Control · Energy Optimization Layer

Problem

Most Indian households lack an affordable, centralized system to manage and optimize their electrical appliances. Existing solutions are either too expensive, fragmented, or lack intelligence — leading to energy waste, manual control, and no real insight into usage patterns.

Approach

Building a cost-effective, AI-assisted home automation system that acts as a central control layer for household devices. The system enables users to monitor, control, and automate appliances while optimizing for energy efficiency and real-world usage behavior.

Vision

To create an accessible smart home infrastructure for Indian households — where automation is not a luxury, but a default layer that improves efficiency, comfort, and energy consumption.

Home Automation IoT Energy Efficiency AI Control
lightbulb Light
ac_unit Climate
videocam Camera
bolt Energy
smartphone App
wifi Network
home AI Hub
Smart Home — AI Control Layer

schemaArchitecture

A central AI hub (SBC/custom PCB) connects to household devices via WiFi and relay modules. The hub runs local inference for usage prediction and scheduling. A mobile app provides real-time control and energy dashboards — no cloud dependency required.

boltTarget

Budget under ₹10,000 for a full household setup. Built for Tier 1 & Tier 2 Indian homes. Expansion path toward small-scale industrial and manufacturing unit automation. Hardware-first, cloud-optional.