# Plooms > One familiar place to use company AI. Your knowledge under your control. Open-weight models and managed model serving. Explore Plooms security, auditability and cost. Plooms is a secure AI workspace for company knowledge and work. The public site says: turn company knowledge into answers, documents and workflows, with control over access and cost. This file restates that site. It does not add customers, ratings, or certifications. Plooms, by Spej. Plooms can be used on its own. Spej can also provide custom integration, implementation and ongoing AI leadership as separately scoped services. Spej is at https://spej.ai. ## Pages - [Plooms | Secure AI for company knowledge and work](https://plooms.ai): One familiar place to use company AI. Your knowledge under your control. Open-weight models and managed model serving. Explore Plooms security, auditability and cost. - [The Platform | Plooms](https://plooms.ai/platform): Plooms combines familiar AI tools with company knowledge, traceable sources and repeatable workflows. Start with the task. Keep the context as you move between capabilities. - [Security & Deployment | Plooms](https://plooms.ai/security): Choose where company knowledge lives, what separates it from other customers and who operates each part of the system. Evaluate those boundaries alongside the policies and handling terms. - [AI for Midmarket & Enterprise | Plooms](https://plooms.ai/enterprise): Company AI with connected knowledge, working documents and approval-based workflows. Explore customer-cloud and fully managed dedicated Plooms deployments. - [AI for Small Business & Nonprofits | Plooms](https://plooms.ai/small-business): The Plooms business workspace on managed shared hosting with logical isolation. Company knowledge, file creation and reusable workflows without a setup fee. - [Personal AI Workspace | Plooms for Individuals](https://plooms.ai/individuals): Connect your knowledge, chat, create documents and build apps in a personal Plooms workspace. A prepaid monthly plan with included usage and no infrastructure setup. - [Resources & About | Plooms](https://plooms.ai/resources): Clear explanations for business leaders, with the architecture and operating detail your technical team needs. - [Talk to the Plooms team | Plooms](https://plooms.ai/contact): Tell us about your users, company knowledge and data requirements. We’ll arrange the right product conversation and discuss deployment and cost. - [Privacy Policy | Plooms](https://plooms.ai/privacy): How Plooms collects, uses, and protects your information, including data from connected Google accounts. - [Terms of Service | Plooms](https://plooms.ai/terms): The terms that govern your use of Plooms. ## What the product includes - Chat with company context - Create documents and media - Build working applications - Explore your company brain - Connect the sources you choose - Run repeatable workflows - Connect your own agents The same company knowledge supports those capabilities. Connectors named on the platform page are Google Workspace, Microsoft 365, Box, Slack and Teams. People choose what Plooms may read. ## Deployment - Small-business hosting uses logical isolation on shared infrastructure. The published shared-business plan uses monthly seats and a prepaid token wallet, with no setup fee or dedicated-infrastructure minimum. Self-serve signups are open at https://plooms.ai/business. - Enterprise options provide dedicated application infrastructure, either managed for you or in your own Azure tenant. Model processing is a separate deployment choice. Requests to run Plooms in your own Azure tenant are open at https://plooms.ai/get-started/byo-managed-gpu. - Managed model serving can be serverless or dedicated, matched to the workload. - Individuals: $25/month, including $10 of AI usage each month. Add prepaid usage when you need more. No setup fee. That price is the personal plan only. ## Compliance SOC 2 Type II: In progress. Report pending · Target completion Q1 2027. HIPAA Business Associate Agreement (BAA): In progress. BAA availability pending. These initiatives are not yet complete. Contact us to discuss current availability and your requirements. Do not describe these as finished or currently available. ## Questions published on the homepage ### Does my company get its own infrastructure? It depends on the plan. Small-business hosting uses logical isolation on shared infrastructure. Enterprise options provide dedicated application infrastructure, either managed for you or in your own Azure tenant. Model processing is a separate deployment choice. ### Can I see where an answer came from? Plooms preserves supporting sources and the history of company knowledge. You can inspect those records to evaluate an answer. Answers still need appropriate review. ### Will Plooms reduce our AI costs? Savings depend on your workload and setup. Compare platform fees, model usage, hosting and setup costs. Plooms gives you visibility and model choice to make that comparison. ### Can I use Plooms without a Spej engagement? Yes. Plooms can be used on its own. Spej can also provide custom integration, implementation and ongoing AI leadership as separately scoped services. ## Questions published on the security page ### Where is our data stored and processed? Application hosting and stored knowledge are configured in your cloud or Plooms hosting. Model processing runs on separately managed model serving, using serverless or dedicated capacity for the workload. The proposed architecture should identify the provider, region and flow of relevant context between them. ### What do open-weight models let us inspect? Open-weight models make the weights available for technical evaluation and deployment. A closed model API does not provide that weight access, though its product can still offer access controls and audit logs. With Plooms, discuss the model, version and update process for your workload. Weight access does not reveal every cause behind an answer or remove the need for validation. ### Does an isolated application mean dedicated GPUs? No. Application and data isolation are separate from model serving. Managed serverless inference can reduce idle-capacity costs; dedicated serving can suit requirements for capacity, latency or stronger compute isolation. We select the serving arrangement for the workload and explain where relevant context is processed. ### What can our team inspect? The workspace supports company context with source provenance and history. Review the specific access controls, event coverage, retention and any export requirements with us for your configuration. ### How do connected sources work? Plooms lets users bring connected knowledge into their workspace. Scope, sharing and administrative policies should be reviewed as part of deployment, including how access changes and deletion are handled. ### Is Plooms automatically compliant for our industry? Compliance depends on your obligations, the configuration and how the system is used. We work through data handling, identity, retention, logging and contractual requirements with your stakeholders before you commit. ### What changes when external tools are used? Web search and connected services can introduce additional data flows. Review which tools are enabled, what information they receive and the applicable handling terms. ### How should we evaluate cost? Start with your workflows, expected demand and capacity requirements. Review setup, seat subscriptions, infrastructure and model usage together, including the agreed operational responsibilities. ## Demo - [Plooms in one minute](https://plooms.ai/#overview-film): See Plooms in action in one minute. Video file: https://plooms.ai/motion/plooms-demo-sept30.mp4. Captions: https://plooms.ai/motion/plooms-demo-sept30.vtt. ## Contact - Email: s@plooms.ai - Contact page: https://plooms.ai/contact ## Do not cite - Pages under /samples/ are illustrative workflow documents labeled as fictional company data. - /resources/temporal-knowledge-graphs, /resources/single-tenant-honest-privacy, and /resources/open-weight-models-in-production are linked from the resources page but are not published articles. They currently return an empty application shell. - /about, /get-started, and /business are not public marketing pages.