Shram Intelligence, Inc.
Shram Intelligence describes itself as the AI cat company behind minimi, building personal intelligence that comes to users and closes their open loops. minimi captures personal context on a Mac, keeps memory locally on-device, and makes it available to AI assistants so users do not repeatedly have to explain themselves or manually manage commitments.
At a Glance
- Mac-based knowledge workers
- Founders and high-agency professionals
- People using Claude, ChatGPT, Gemini, or other LLMs who want persistent personal context
- Teams that need shared context while retaining personal data ownership
- +1 more
AI Tools by Shram Intelligence, Inc.
(1)
minimi
Mac AI Memory Layer for Devs
Discussions
No discussions yet
Be the first to start a discussion about Shram Intelligence, Inc.
Latest News
minimi 2.0 launched on Product Hunt with Melody, an AI cat for finding and closing open loops; ranked #4 of the day.
Jay Gadekar and Ojasvika Sahu discussed Shram and minimi's origin, architecture, privacy model, user growth, and plans for more fully on-device memory on The Offline Network.
Minimi launched publicly as a Mac-based ambient memory layer that connects personal context to any LLM through MCP.
Organize by Shram Chrome extension updated to version 1.0.3; it organizes browser tabs with local, privacy-first AI.
Products & Services
A Mac app and MCP-based personal memory layer that captures what the user reads, says, and hears, stores memories locally, and makes them available to Claude, ChatGPT, Gemini, and other LLMs. Cotton is the AI cat/persona associated with personal context and memory.
The second AI cat in minimi 2.0. Melody uses the user's personal context to find open loops and close commitments automatically, without manual prompting or typing context.
The company's earlier AI-native work-management/productivity product, designed to connect daily work with performance, project tracking, real-time progress, tasks, and follow-up work.
A Chrome extension that automatically organizes open browser tabs into logical groups using local, privacy-first, on-device AI.
Market Position
Shram positions minimi as an ambient, invisible memory layer rather than another manually maintained notes database or an LLM-specific memory silo. Its differentiators are Mac-level capture, local storage, MCP interoperability across LLMs, and automatic open-loop closure. The founders explicitly compared use cases with Granola and Obsidian and positioned minimi against the blind spots of native LLM memory and manual second-brain tools.
Leadership
Founders
Jay Gadekar
Co-founder and CEO of Shram/minimi. Public profiles describe him as building minimi and previously working as a self-employed builder/tinkerer before the company; he leads the product and company direction.
Ojasvika Sahu
Co-founder and Chief Design Officer of Shram/minimi. She is publicly described as a product designer who crafts AI products and is responsible for design and user experience.
Vineet Gupta
Co-founder and the founding ML/AI researcher behind the core memory technology. Public search results describe him as the person behind the core tech and as a founding AI researcher at Shram.
Executive Team
Jay Gadekar
Co-founder and CEO
Leads Shram/minimi and has publicly discussed the product architecture, product strategy, and company direction.
Ojasvika Sahu
Co-founder and Chief Design Officer
Leads design and user experience; she has publicly demonstrated minimi use cases and described the product's privacy-first interaction model.
Founding Story
Shram began as a bootstrapped productivity/work-management company. After users repeatedly asked to access the ambient capture and on-device memory layer used inside Shram as a personal second brain, the team made minimi as a side project. The project was also inspired by the death of the founders' kitten, Cotton: the team built an early version in three days, named it 'mini-me,' and made Cotton the mascot. The product was then developed into a Mac app that captures context locally and connects it to an LLM through MCP.
Business Model
Revenue Model
Freemium software subscription. The basic minimi experience is free, with a higher tier intended for users who need more MCP calls; the company was still testing the conversion limit in July 2026.
Pricing Tiers
The company said the free tier provided up to five MCP calls per week, described in the interview as roughly once per day.
For users needing more than the free MCP-call allowance.
Target Markets
- Mac-based knowledge workers
- Founders and high-agency professionals
- People using Claude, ChatGPT, Gemini, or other LLMs who want persistent personal context
- Teams that need shared context while retaining personal data ownership
- Users seeking privacy-first, local AI productivity tools
- Personal second-brain and long-term memory for AI chats
- Finding forgotten commitments, action items, and follow-ups
- Meeting recall and meeting-note retrieval without a meeting bot
- Searching across email, messaging, browser reading, and offline conversations
- Connecting personal context to Claude or another LLM through MCP
- Knowledge workers, founders, and high-agency professionals who manage many information streams
- The company reported approximately 1,000 users in more than 20 countries in July 2026.
- A 73-year-old former NASA engineer identified as Mr. Bill was described by the founders as an enthusiastic user.
- A user identified as Charles was described as moving from another memory product after finding minimi's retrieval substantially faster.