--- title: The Factory That Dreams: 39 AI Agents, No Framework -... description: Rushabh Doshi, founder of Machinecraft, a 100-person manufacturing company in India, shares how he built a 36+ AI agent system called Eira to preserve and... canonical: https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft | daily.dev og:description: Rushabh Doshi, founder of Machinecraft, a 100-person manufacturing company in India, shares how he built a 36+ AI agent system called Eira to preserve and... og:url: https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi og:image: https://api.daily.dev/og/posts/OJn0T2RYi.png og:image:alt: The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft **[AI Engineer](https://daily.dev/sources/aidotengineer)** · 9 min read · 2 upvotes · 0 comments ## Summary Rushabh Doshi, founder of Machinecraft, a 100-person manufacturing company in India, shares how he built a 36+ AI agent system called Eira to preserve and operationalize decades of institutional knowledge — with no data science team and no model training. The system uses RAG with vector and graph databases, a layered memory architecture (working memory, episodic memory, relationship warmth), and a nightly 'dream cycle' that consolidates knowledge. Agents are organized as specialists (Athena, Prometheus, Plutus, etc.) that collaborate to handle the entire go-to-market workflow: outbound emails, quotations, lead revival, and inbound qualification. Built for ~$30K instead of a quoted $230K, it runs on a few thousand dollars per month. The architecture has been extracted into an open, forkable template called Brain OS at forkmybrain.org. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How can a small company build a multi-agent AI system without a data science team or training its own model? By feeding all internal documents into off-the-shelf language models to extract facts stored as vectors and relationship graphs, rather than fine-tuning or training a custom model. One factory built a 36-agent system this way, each agent specialized for one task (pricing, sales, fact-checking), coordinated through a single interface exposing over 200 tools via one protocol, for around $30,000 total and a few thousand dollars monthly running cost. _See how developers on daily.dev track approaches to building multi-agent systems without a training budget._ ### What is a nightly dream cycle in an AI agent memory system? A dream cycle is a nightly process where an AI system replays the day's interactions, consolidates useful information, detects contradictions between facts, discards stale data, and converts the day's work into reusable skills. It produces a summary report each morning describing what was consolidated, what was discarded, and what new patterns were identified, effectively making the system more capable overnight without retraining. _Developers designing agent memory layers follow discussions like this on daily.dev._ ## Similar posts on daily.dev - [My Agent Setup](https://daily.dev/posts/my-agent-setup-8ttve5oux) · Hacker News · 1 upvotes · 0 comments - [This startup wants to make enterprise software look more like a prompt](https://daily.dev/posts/this-startup-wants-to-make-enterprise-software-look-more-like-a-prompt-t1uuynped) · TechCrunch · 1 upvotes · 0 comments - [Engineering Multi-Agent AI Teams That Build and Test Themselves](https://daily.dev/posts/engineering-multi-agent-ai-teams-that-build-and-test-themselves-ehcixev11) · Salesforce Engineering · 1 upvotes · 0 comments - [Building Reliable Agentic AI Systems](https://daily.dev/posts/building-reliable-agentic-ai-systems-jvjrmw6pg) · Martin Fowler · 4 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#productivity](https://daily.dev/tags/productivity), [#rag](https://daily.dev/tags/rag), [#vector-search](https://daily.dev/tags/vector-search) [View this post on daily.dev](https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft","url":"https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi"},"datePublished":"2026-07-11T20:03:30.106Z","dateModified":"2026-09-14T08:31:26.342Z","description":"Rushabh Doshi, founder of Machinecraft, a 100-person manufacturing company in India, shares how he built a 36+ AI agent system called Eira to preserve and...","image":"https://i.ytimg.com/vi/jtzh-GBXBWc/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/jtzh-GBXBWc/sddefault.jpg","isAccessibleForFree":true,"articleSection":"AI Engineer","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"AI Engineer","logo":"https://media.daily.dev/image/upload/s--u5PucxNT--/f_auto/v1724338940/logos/aidotengineer","url":"https://daily.dev/sources/aidotengineer"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,productivity,rag,vector-search","timeRequired":"PT9M","video":{"@type":"VideoObject","name":"The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft","description":"Rushabh Doshi, founder of Machinecraft, a 100-person manufacturing company in India, shares how he built a 36+ AI agent system called Eira to preserve and...","thumbnailUrl":"https://i.ytimg.com/vi/jtzh-GBXBWc/sddefault.jpg","uploadDate":"2026-07-11T20:03:30.106Z","duration":"PT9M","url":"https://api.daily.dev/r/OJn0T2RYi","embedUrl":"https://www.youtube.com/embed/jtzh-GBXBWc"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"AI Engineer","item":"https://daily.dev/sources/aidotengineer"},{"@type":"ListItem","position":3,"name":"The Factory That Dreams: 39 AI Agents, No Framework - Rushabh Doshi, Machinecraft"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-factory-that-dreams-39-ai-agents-no-framework---rushabh-doshi-machinecraft-ojn0t2ryi#faq","mainEntity":[{"@type":"Question","name":"How can a small company build a multi-agent AI system without a data science team or training its own model?","acceptedAnswer":{"@type":"Answer","text":"By feeding all internal documents into off-the-shelf language models to extract facts stored as vectors and relationship graphs, rather than fine-tuning or training a custom model. One factory built a 36-agent system this way, each agent specialized for one task (pricing, sales, fact-checking), coordinated through a single interface exposing over 200 tools via one protocol, for around $30,000 total and a few thousand dollars monthly running cost. See how developers on daily.dev track approaches to building multi-agent systems without a training budget."}},{"@type":"Question","name":"What is a nightly dream cycle in an AI agent memory system?","acceptedAnswer":{"@type":"Answer","text":"A dream cycle is a nightly process where an AI system replays the day's interactions, consolidates useful information, detects contradictions between facts, discards stale data, and converts the day's work into reusable skills. It produces a summary report each morning describing what was consolidated, what was discarded, and what new patterns were identified, effectively making the system more capable overnight without retraining. Developers designing agent memory layers follow discussions like this on daily.dev."}}]} ```