Kakoli

Kakoli is a physical AI robotic chef, an induction cooker and 3D-printer-based arm controlled via a Flutter app, that learns your recipes by demonstration and cooks them on demand.

Category

Physical AI

Robotics

Client

Internal

Website

https://youtube.com/shorts/k8B3WlDiTCQ

Date

Kakoli is a physical AI project: a robotic chef that pairs a RaspberryPi-controlled induction cooker with a 3D-printer-derived arm to pick, drop, and stir ingredients, letting it execute recipes with the same precision every time. Users teach it recipes by demonstration through a Flutter mobile app (built for iOS, Android, iPad, and web from one codebase), then just load ingredients and hit start while Kakoli follows the recorded steps, tracks kitchen inventory, and recommends meals from what's on hand. The stack spans embedded control (Python-based path finding and motor control on Linux, Marlin firmware for the actuator) up through a mobile app and backend SaaS platform, aimed at making healthy, home-cooked meals accessible to people with limited time, like working parents and elderly users.

More projects

Zyro

Zyro is a US-based startup and participant in the Antler VC Residency Program, building a "trust layer" for the internet where user preferences and influence are verified and fairly compensated rather than silently scraped.

Client

Zyro, US

Date

May 1, 2025

Zyro

Zyro is a US-based startup and participant in the Antler VC Residency Program, building a "trust layer" for the internet where user preferences and influence are verified and fairly compensated rather than silently scraped.

Client

Zyro, US

Date

May 1, 2025

Zyro

Zyro is a US-based startup and participant in the Antler VC Residency Program, building a "trust layer" for the internet where user preferences and influence are verified and fairly compensated rather than silently scraped.

Client

Zyro, US

Date

May 1, 2025

Mission Control

A self-hosted AI agent team (orchestrator + 4 specialists) with a Next.js/SQLite dashboard for tracking projects and tasks, packaged as a one-command Docker deploy.

Client

Prospect Client

Date

Apr 4, 2025

Mission Control

A self-hosted AI agent team (orchestrator + 4 specialists) with a Next.js/SQLite dashboard for tracking projects and tasks, packaged as a one-command Docker deploy.

Client

Prospect Client

Date

Apr 4, 2025

Mission Control

A self-hosted AI agent team (orchestrator + 4 specialists) with a Next.js/SQLite dashboard for tracking projects and tasks, packaged as a one-command Docker deploy.

Client

Prospect Client

Date

Apr 4, 2025