Client loginBook a call
Build
Custom Software DevelopmentSystems built to your process
Web Application DevelopmentReact, Next.js, Rails, Django
Mobile App DevelopmentiOS, Android, React Native
App DevelopmentWeb and mobile apps, end to end
SaaS Product DevelopmentMulti-tenant from day one
MVP DevelopmentFirst release in 6–10 weeks
Enterprise SoftwareSSO, audit trails, scale
Connect & operate
API Development & IntegrationREST, GraphQL, webhooks
Cloud & DevOps EngineeringAWS, GCP, Azure, Terraform
QA & Software TestingAutomated suites, load, E2E
Legacy ModernizationStrangler-fig migrations
Cybersecurity EngineeringHardening and remediation
Data EngineeringPipelines, warehouses, reporting
Design & teams
UI/UX DesignInterfaces people finish tasks in
Prototyping & DiscoveryScope before you spend
Staff AugmentationSenior engineers on your team
Dedicated Dev TeamsA full pod, one contract
White-Label DevelopmentEngineering under your brand
Technology ArchitectureDecide once, correctly
Project ManagementDelivery you can audit
Start hereScope a build in one callBring a spec, a wireframe, or a paragraph. You leave with an architecture, a timeline, and a number.Book a scoping call
All software development services →How we price work →Talk to an engineer →
AI software
AI Application DevelopmentFull-stack AI products
AI Agent DevelopmentMulti-step autonomous agents
AI SaaS DevelopmentMulti-tenant AI products
Workflow AutomationReplace manual ops
Custom LLM AppsBranded internal assistants
LLM & data systems
Private LLM DevelopmentSelf-hosted Llama, Mistral, Qwen
RAG DevelopmentCited answers over your docs
MCP & WebMCPMake your systems agent-callable
Vector Database Setuppgvector, Pinecone, Qdrant
Document IntelligenceOCR and extraction pipelines
Vibe coding
AI-Assisted BuildsFast, with senior review
Rapid PrototypingClickable in days
Concept to CodeIdea to working software
Code Review & CleanupMake AI output shippable
Fractional AI TeamOngoing build capacity
Ready to ship?Put AI where the work isAgents, RAG, and private LLMs wired into the systems your team already uses — not a chatbot bolted to a homepage.Discuss an AI project
AI development services →AI readiness checklist →AI vs. traditional development →
Domain firstWe learn your workflow before we model itRegulated, operational, or high-volume — the constraints belong in the schema, not in a training doc.Talk about your domain
Selected workSoftware we own and operateMost agencies show screenshots. We run the products — the same stack, uptime, and on-call we hand to clients.View the portfolio
Plan smarterEstimate before you commitCost ranges, scope templates, and the questions we ask in discovery — free, no form.Open the cost calculator
Real conversationsTalk with a technical leadNo SDR, no discovery gauntlet. The person on the call is the one who scopes the build.Book a call
RAG Frameworks · mindsdb

minds

MindsHub Cowork is an open-source AI agent platform that automates multi-step knowledge work tasks—reports, analysis, monitoring—by routing them to the right LLM and turning outputs into shareable artifacts. It runs on your machine, your cloud, or their hosted service, with built-in connections to data systems like BigQuery, Postgres, and Notion.

Talk to DEV.coDiscuss Implementation
Source: GitHub — github.com/mindsdb/minds
39.4k
GitHub stars
6.2k
Forks
Makefile
Primary language
MIT
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositorymindsdb/minds
Ownermindsdb
Primary languageMakefile
LicenseMIT — OSI-approved
Stars39.4k
Forks6.2k
Open issues3
Latest releasev26.1.0 (2026-04-23)
Last updated2026-07-01
Sourcehttps://github.com/mindsdb/minds

What minds is

A superproject pulling together a desktop/web frontend (Electron and React SPA), a Python agent backend supporting open models and frontier APIs, and a data vault layer. Built with Makefile automation; supports Python 3.10–3.13; uses submodule-based development workflow with hot-reload modes for both desktop and web.

Quickstart

Get the minds source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/mindsdb/minds.gitcd minds# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Internal BI & reporting automation

Connect data warehouses and automate recurring report generation, analysis, and distribution without building custom dashboards or pipelines.

Multi-step research and content workflows

Delegate research tasks, competitive analysis, or content drafting to agents; collect finished artifacts as docs, decks, or code ready to publish.

Scheduled monitoring and alert tasks

Run cross-system monitoring jobs on a schedule—pull data from multiple sources, analyze, and generate alerts or summaries without custom orchestration.

Implementation considerations

  • Data vault connection setup—each connection (BigQuery, Postgres, etc.) requires credentials scoped per agent; plan credential rotation and audit logging.
  • Model provider keys—supports Claude, GPT, Gemini, DeepSeek, Qwen, and others; cost and latency vary by router configuration and task complexity.
  • Submodule-based development requires discipline: use `dev.env` and `make pin` to avoid merge conflicts when working on feature branches across frontend/backend/data-vault.
  • Fresh deployments run `make flush` to reset state; plan data migration if moving between instances (conversations, keys, projects stored in `~/.anton` and `~/.cowork`).
  • Memory and skill persistence—cross-session state depends on local database and skill library maintenance; document retention policies upfront.

When to avoid it — and what to weigh

  • Need strict real-time performance guarantees — Agent-based workflows involve LLM latency and multi-step processing; not suitable for sub-second or hard real-time requirements.
  • Require guaranteed off-the-shelf compliance certifications — MIT license and open-source nature mean security/compliance posture depends on your deployment, audit, and hardening—no pre-packaged SOC 2 or FedRAMP claims.
  • Locked into proprietary BI or workflow tools — Cowork replaces some traditional BI and workflow automation; switching requires re-architecting existing pipelines and integrations.
  • Very low operational maturity or DevOps capacity — Self-hosting from source requires Makefile familiarity, Python environment management, and debugging multi-service deployments; hosted option mitigates this.

License & commercial use

MIT License. Permissive OSI license allowing commercial use, modification, and distribution with minimal restrictions. Bundled submodules have their own licenses—review each submodule repository (frontend, backend/core_api, backend/core_agent, backend/data-vault) for their terms.

MIT permits commercial deployment and derivative works. However, no indemnity or liability cap is provided by the license. For production use, assess: (1) your own liability if agents misbehave or data is breached; (2) support model (community Discord vs. enterprise SLA contact); (3) liability for third-party integrations and model providers (Claude, OpenAI, etc. have their own terms). Recommend legal review before large-scale commercial deployment.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationAdequate
License clarityClear
Deployment complexityHigh
DEV.co fitGood
Assessment confidenceHigh
Security considerations

Credential management: vault-scoped per connection, but security depends on your deployment (who has access to ~/.anton, ~/.cowork). LLM model routing introduces trust in third-party API endpoints (Claude, OpenAI, etc.). Open-source code is auditable but provides no pre-packaged hardening, TLS enforcement, or audit logging. Private security reporting available via GitHub security policy; no mention of penetration testing, threat model, or incident response SLA. Air-gapped deployments supported per README, but vetting custom model endpoints is your responsibility.

Alternatives to consider

Retool

Mature internal-tool and workflow builder with UI-first design; better for teams less comfortable with code. Less emphasis on agentic delegation; more on CRUD and form automation.

n8n

Workflow automation and integration platform; visual node editor, larger ecosystem of pre-built connectors. Oriented toward ETL and webhook-driven tasks, not agentic research or multi-turn reasoning.

Zapier / Make

Cloud-native workflow automation with massive third-party integration library. Less suited to complex reasoning tasks; strong for simple trigger-action and multi-app orchestration.

Software development agency

Build on minds with DEV.co software developers

Start with the hosted web app at console.mindshub.ai, or build from source. Join the Discord community for support and feature discussion.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

CowAgentpipeshub-aidbhub

Related on DEV.co

Explore the category and the services that help you build with it.

RAG FrameworksAI Development ServicesAI DevelopmentAPI DevelopmentCloud DeploymentTalk to DEV.co

minds FAQ

Can I run this entirely air-gapped (no cloud calls)?
Yes, per README. You can self-host with open models (DeepSeek, Qwen, Kimi) and local data connections. You must deploy model endpoints yourself and route agents to them; frontier model calls (Claude, GPT) cannot be air-gapped without proxy/relay.
What happens to my conversations, keys, and projects if I switch from desktop to web or vice versa?
Not clearly stated in the excerpt. Desktop stores state in ~/.anton and ~/.cowork; web app uses their hosted backend. Migration between modes likely requires manual export/import or sync setup. Verify with docs or contact support before switching.
Do I need to provision GPU or custom compute for agents?
Depends on your model choice. Frontier models (Claude, GPT) use their cloud; open models (DeepSeek, Qwen) can run on modest CPU if you self-host or use a third-party endpoint. No GPU requirement mentioned; inference latency and cost vary by provider.
Can I integrate custom data sources or model endpoints?
Data vault supports BigQuery, Postgres, MySQL, MSSQL, Gmail, Drive, HubSpot, Notion, Linear. Model router supports Claude, GPT, Gemini, DeepSeek, Qwen, Kimi. Custom integrations require code contribution to the repo or running a proxy/adapter layer.

Senior engineers for your next build

Our engineers ship rag frameworks software for a living. DEV.co provides software development services, web development services, and ongoing support for teams standardizing on tools like minds.

Ready to automate your knowledge work?

Start with the hosted web app at console.mindshub.ai, or build from source. Join the Discord community for support and feature discussion.

Talk to DEV.coSchedule a Consultation