Built with passion • Open Source • Community-driven

About Askimo

A personal project to build the AI tool I wished existed

The Origin Story

Why I Built Askimo

When I first started working on Askimo, I didn't plan to build a big project. I just wanted something simple that could solve the problems I was facing every day.

The Problems I Faced Daily

Like many people trying to get the most out of AI tools, I was juggling multiple applications and services. Some tools solved one or two of these problems, but none of them solved all of them:

Browser-based tools that crashed during long conversations

Most AI clients are just web browser wrappers. After a lengthy chat session, the browser would freeze, and I'd lose my entire conversation history. (Some native apps existed, but they locked you into one AI provider.)

Repeatedly explaining my context and data

Every new chat session meant re-introducing my project details, my files, and my preferences. (RAG tools existed, but they didn't support multiple AI providers or had poor UX.)

No customization for my writing style and needs

I wanted to define my tone, my preferences, and switch between them quickly. (Tools with good customization often lacked other critical features.)

Network restrictions and availability issues

At home I could use OpenAI. At work they were blocked. (Local AI solutions existed but switching between local and cloud was complicated.)

Manual, repetitive tasks

I wanted AI to help summarize changes and automate routine work. Instead I found myself doing the same things over and over. (Automation tools didn't integrate well with desktop apps.)

The Moment It Clicked

What if I could build a native AI tool that solved all of these problems in one place?

Not just another web wrapper. Not just another single-provider app. A real desktop application that brings everything together:

Runs natively without browser performance issues
Remembers your context automatically with RAG
Switches between AI providers seamlessly
Works offline with local models when needed
Automates repetitive workflows to save time

That idea became the starting point for Askimo.

The Vision

The Bigger Vision

Askimo started as a tool to solve my own problems, but it represents something bigger:

Empowering individuals and small teams to leverage AI without enterprise resources.

Right now, only big corporations have the resources to develop custom AI solutions. They build dedicated teams, proprietary RAG systems, and tailored decision-support agents.

But what about:

  • 🧑‍💼 The solo developer who wants AI to understand their entire codebase?
  • 💼 The small business owner who needs AI to analyze their data?
  • 📈 The individual investor who wants to customize which sources AI uses for stock research?
  • 🏢 The startup that can't afford a dedicated AI team?

Askimo aims to democratize AI tooling. With RAG indexing, smart decision-support agents, and multi-provider flexibility, anyone can build their own AI-powered workflows.

What Makes Askimo Different

Built for Real Use

Askimo isn't a side project I built once and forgot about. It's something I use every day. Every feature exists because it solved a real problem.

Privacy-First

All your data stays local. RAG indexes are stored on your machine. API keys are encrypted. No telemetry, no cloud sync.

Learn about our security & privacy

No Vendor Lock-In

Switch between OpenAI, Claude, Gemini, Ollama, and 8+ providers without changing tools. Your chat history and projects stay with you.

Open Source

AGPLv3 licensed. Anyone can use it, modify it, or contribute to it. Transparency and community-driven development.

Where We Are Now

Askimo is actively developed and already provides significant value:

Core Features

Desktop app for macOS, Windows, Linux
RAG indexing with Lucene + JVector
CLI automation for DevOps
Project-based sessions
Searchable chat history
Multi-language UI (9 languages)

Supported AI Providers (8+)

Cloud Providers

  • OpenAI Supports the latest OpenAI models
  • Anthropic Claude Claude models (Sonnet, Opus, Haiku)
  • Google Gemini Gemini models (Flash, Pro, Ultra)
  • Grok Grok models (standard and mini)

Local & Self-Hosted

  • Ollama Run any local model (Llama, Mistral, etc.)
  • LM Studio Local model server
  • LocalAI OpenAI-compatible self-hosted
  • Docker AI Containerized AI models

What's Coming Next

The roadmap ahead is focused on making AI more accessible and powerful for everyone:

Enhanced RAG capabilities

Ingest knowledge from diverse sources: DB schemas, SEC filings, market data, academic papers, competitor analysis, e-commerce catalogs, and more.

Smart decision-support agents

Investment advisors, code reviewers, business analysts, research assistants, legal advisors, and product managers — all customizable to your domain.

Have an idea for a new feature?

Askimo's roadmap is shaped by the community. Submit your ideas on GitHub and help shape the future of Askimo.

Submit a Feature Request

About the Creator

Hai Nguyen - Creator and Lead Developer of Askimo

Hai Nguyen

Creator & Lead Developer

I'm someone who builds tools to help people utilize AI better and more flexibly. As Askimo's first user, I started with my own needs, but the tool has evolved based on feedback from users, teammates, and the community.

Askimo is constantly evolving alongside AI advancements. It's taught me that AI doesn't replace work — it extends it.

Open Source & Community

Askimo is 100% open source (AGPLv3 license). This means:

  • ✅ Free to use for personal and commercial projects
  • ✅ Transparent development on GitHub
  • ✅ Community-driven feature requests and bug fixes
  • ✅ No hidden costs or proprietary lock-in
  • ✅ If you modify and distribute the software, you must share your changes

Want to Contribute?

Whether you're reporting bugs, suggesting features, writing documentation, or contributing code — all contributions are welcome!

Get in Touch

Get Started Free

Ready to Try Askimo?

Join the community building the future of developer-friendly AI tools. Free, open source, and built for real use.

Free & open source · No account required · Works offline with local models