Frequently Asked
Questions.
Have questions about security, integration sandboxes, vector memories, or custom cron schedules? Find answers below.
What is Remii?
Remii is a 24/7 virtual assistant and AI employee. Unlike standard chat assistants, Remii is designed to actively operate in the background—completing tickets, scraping websites, syncing spreadsheets, and updating CRMs autonomously.
How do I get started with Remii?
You can sign up for an account via our Web Dashboard, complete the short onboarding profile to set your preferences, and start interacting with your assistant immediately.
Is there a desktop app for Remii?
Currently, Remii is accessible through our fully responsive web application and Telegram bot. Slack, WhatsApp, and Discord integrations are coming soon. A dedicated desktop wrapper is also on our roadmap.
Can I run multiple Remii assistants?
Yes, our Mini, Me, and Us tiers allow you to configure and run multiple specialized AI agents with distinct instructions and connected toolkits.
What makes Remii different from standard custom GPTs?
Standard custom GPTs are reactive, conversational bots. Remii is proactive and agentic: it operates in secure sandboxes, runs background cron schedules, recovers from failures, and handles multi-step tool workflows autonomously.
Do I need coding experience to use Remii?
No. While Remii is built to assist with developer operations, it is controlled entirely via natural language. You can describe your instructions in plain English (or other supported languages).
Can I use Remii for personal tasks?
Absolutely. Remii is great for automating personal routines like summarizing news, managing calendar invites, scraping listings, or sending alerts.
Does Remii support voice commands?
Currently, we support text commands on both the web dashboard and Telegram. Voice-to-text integration is under active development.
What is an 'AI employee'?
An AI employee is an autonomous agent capable of performing discrete digital jobs—such as operations, support, coding, or lead generation—working continuously with minimal supervision.
What is Sandboxed Execution?
Every script, tool action, or terminal command run by Remii occurs inside an isolated, temporary remote container sandbox. This strict boundary prevents any potential prompt injections or malicious scripts from accessing your physical device, local networks, or private files.
How does Remii connect to my external applications?
Remii connects securely using OAuth 2.0 via Composio. We never request, see, or store your raw application passwords, database credentials, or secret keys. All access tokens are fully encrypted and only used in isolated tool execution calls.
Are my application credentials encrypted?
Yes. All authorization keys and OAuth tokens are encrypted at-rest using industry-standard AES-256 GCM encryption before being saved in our secure database layer.
How long do the sandbox containers live?
Sandbox containers are short-lived. They are dynamically spun up for each execution run and completely destroyed upon completion or timeout, leaving no persistent data behind.
Does Remii support Multi-Factor Authentication (MFA)?
Yes, our web dashboard supports and recommends Multi-Factor Authentication (MFA) to protect your account and connected integrations.
Is your platform SOC 2 certified?
We are currently in the audit window for SOC 2 Type II compliance. We adhere strictly to SOC 2 security protocols for data encryption, isolation, and access logs.
What database security measures do you use?
Our PostgreSQL database uses strict Row-Level Security (RLS) policies, ensuring no user can ever read, update, or compromise another user's vector embeddings or settings.
How do you protect against prompt injection attacks?
We employ multiple defense-in-depth measures: input preprocessing filters, separate system prompt isolation, and most importantly, isolated sandboxed runtimes that restrict any injected commands from doing harm.
Can I inspect the raw commands Remii runs?
Yes. Full audit logs are available in your web dashboard. You can inspect the exact code generated, terminal commands initiated, and API payloads sent by your agent.
Do you encrypt data in transit?
Yes, all data transmitted between your browser, our API endpoints, and our background queue workers is encrypted using TLS 1.3.
Do you sell or monetize my private data?
No. Your data is strictly yours. We do not inspect, cache, sell, or train public models on your private conversations, memory indexes, or documents. You have full authority to reset your vector memories or delete your account instantly at any time.
How can I delete my data?
You can delete all conversation history, vector memories, and active tool connections directly in the account settings page. Account deletion is permanent and wipes all associated records instantly.
Is Remii GDPR compliant?
Yes. We fully support GDPR mandates, including the right to be forgotten, the right to data portability, and the right to object to data processing.
Are conversations sent to LLMs used to train their models?
We use zero-data-retention APIs for Anthropic, OpenAI, and DeepSeek. Under these agreements, our provider partners do not use your inputs or outputs to train their public models.
Where is my data hosted?
Our databases and application servers are hosted in secure, isolated regions inside Amazon Web Services (AWS) and Google Cloud Platform (GCP).
Who has access to my vector memory table?
Only your specific authenticated agent has access. It is queries at runtime using strict Row-Level Security policies to supply context to the LLM.
Does Remii use trackers or third-party cookies?
We only use essential functional session cookies to manage authentication. We do not use third-party advertising trackers or behavioral analytics pixels.
What is your data retention policy for logs?
Standard execution logs are retained for 30 days to facilitate debugging and auditing, after which they are permanently deleted.
Can I export my chat logs?
Yes. You can request a full JSON export of your complete interaction history and vector memories at any time from your settings panel.
Which applications can Remii connect to?
Remii integrates with over 1,000 tools including Slack, Gmail, GitHub, Linear, Google Sheets, HubSpot, Notion, Stripe, Discord, and Shopify.
Does Remii support Slack?
Slack support is coming soon! We are actively building a dedicated Slack app so you can message your agent and delegate tasks directly within your workspace channels. WhatsApp and Discord are also on the roadmap.
Does Remii support Telegram?
Yes! Telegram is live and fully supported. Connect your profile to a Telegram bot token and you can run background actions, query metrics, and receive notifications on the go. WhatsApp, Slack, and Discord channels are coming soon.
How many integrations can I connect at once?
The number of active tool connections depends on your subscription tier: up to 1 for the Free plan, up to 3 for the Mini plan, up to 20 for the Me plan, and up to 40 active connections for the You plan.
Can Remii write code and push commits?
Yes. Given the appropriate OAuth permissions, Remii can pull repository code, edit files, test changes in its sandbox, compile release logs, and push branches or pull requests.
Can Remii send emails on my behalf?
Yes, you can link your Gmail or Outlook account via secure OAuth. You can then instruct Remii to write drafts, parse support queues, or send notifications.
Can Remii manage my calendar?
Yes. By linking Google Calendar, Remii can schedule meetings, check availability, reschedule clashes, and summarize daily plans.
Can I connect local files or databases?
Yes. You can upload CSVs, text documents, or PDFs. For local databases, you can securely route connection requests through an SSH tunnel in the sandbox.
What is Composio?
Composio is our integration backend partner. It provides audited toolboxes, secure token management, and reliable API triggers to connect LLMs to external web apps.
Can I schedule recurring automation workflows?
Yes. You can instruct Remii to perform actions on a recurring cron schedule, such as 'Every Monday at 9 AM, pull GitHub release notes and compose a newsletter draft in Notion'. The state machine runs continuously in the background to handle these schedules.
Can Remii loop indefinitely?
To prevent runaway loops and billing shocks, all execution runs have default timeouts and loop count limits (e.g., maximum 30 iterations or 15 minutes per task).
Can I set custom trigger events?
Yes. You can set Webhook triggers. When an external system (like Stripe or Shopify) triggers a webhook, it wakes your agent to execute a pre-defined instruction.
Can Remii monitor websites for changes?
Yes. You can schedule Remii to scrape a target URL periodically, compare the raw text to previous vector memories, and alert you if specific changes are detected.
How does Remii handle CAPTCHAs?
Remii executes commands programmatically. If a target website serves a CAPTCHA, the agent will pause and log a notification requesting human intervention.
Can I trigger workflows from Slack messages?
Yes. By mentioning your agent in Slack (e.g., '@remii generate report'), you can start background runs immediately without logging into the dashboard.
Is there a limit to how many schedules I can run?
The number of active cron schedules depends on your subscription plan: Free supports up to 3 active cron jobs, Mini up to 7, Me up to 50, and the Us plan supports up to 250 active scheduled tasks.
Can I pause and resume active schedules?
Yes, you can toggle schedules on or off instantly from the automation dashboard panel.
Does Remii support conditional branching?
Yes. Because workflows are orchestrated by LLM reasoning, you can define logic like: 'If transaction is greater than $100, alert Slack, otherwise log it quietly in Notion'.
How does long-term vector memory work?
As you chat with Remii, it automatically extracts relevant workspace context, onboarding preferences, and execution rules. These are converted to semantic vector memory embeddings. When starting a new task, Remii queries these embeddings to retrieve background context instantly.
What is a vector embedding?
A vector embedding is a mathematical representation of text meaning. It allows Remii to perform semantic searches, matching your queries based on intent and topic rather than exact keywords.
Can I manually add items to memory?
Yes. You can instruct Remii directly: 'Remember that our staging domain is staging.remii.space'. It will save this instruction in your persistent vector context.
How do I clear or reset Remii's memory?
You can search, edit, or wipe specific memory records, or perform a complete memory table factory reset from the web dashboard.
Does memory persist across different channels?
Yes. The vector memory table is centralized. What Remii learns on the web dashboard carries over when you chat with it on Telegram. When WhatsApp, Slack, and Discord launch, they will share the same persistent memory.
How large can the memory bank grow?
The Free tier supports standard memory storage, the Mini and Me tiers include automatic memory extraction, and the Us tier features durable vector memory search with high-capacity indexing.
Does Remii memorize passwords?
No. Our onboarding filters actively prevent the agent from extracting or saving plaintext passwords or keys into the vector memory table.
Can I upload a document to use as memory?
Yes. You can upload PDFs, text files, or markdown guides. Remii will chunk the text, generate embeddings, and load them into your context table.
Which LLM models does Remii support?
Remii supports leading model providers, including Anthropic (Claude), OpenAI (GPT-4o), and DeepSeek. You can swap models, update your system instructions, and customize the reasoning depth directly through your dashboard settings.
How does Remii choose which tools to run?
Remii uses tool-calling models (Function Calling). By matching your request to the description schemas of your connected integrations, the model determines the precise API payload.
What is context window limitation?
The context window is the memory capacity of the LLM in a single turn. Remii uses smart retrieval-augmented generation (RAG) to only inject the most relevant memory records, avoiding window overflow.
Can Remii output JSON structure?
Yes. By utilizing structured outputs, Remii can guarantee that tool payloads or reports adhere to precise schemas.
How do you handle model hallucinations?
We use structured tool validations and multi-phase validation loops. If a model tries to invoke a tool with invalid inputs, our execution runtime returns the error description, guiding it to correct itself.
Can I adjust the agent's creativity (temperature)?
Yes. You can adjust the reasoning temperature in your settings: lower for precise task execution, and higher for creative copywriting.
Does Remii support multi-lingual chats?
Yes. The underlying models understand and respond in over 80 languages, allowing you to prompt in Spanish, Japanese, German, etc.
How fast are response generations?
Simple tool actions complete in 1-3 seconds. Multi-step reasoning loops that scrape sites and build logs can take 10-30 seconds depending on API speeds.
Is there a free trial?
Yes, our Free Tier gives you a one-time $1.00 free credit to test integrations, connect toolkits, and build memories.
Do you support local currencies?
Yes! We support multiple local currencies, including INR pricing with convenient direct payment options like UPI for developers in India.
What are 'execution credits'?
An execution credit is consumed when Remii successfully invokes an external tool or runs code inside its sandbox container.
Can I cancel my subscription anytime?
Yes. You can cancel or change your plan at any point. Your premium access will continue until the end of the current billing cycle.
Do credits roll over to the next month?
No. Unused credits from your monthly allocation expire at the end of the billing period and do not roll over.
How does billing handle failed runs?
If a workflow fails due to an infrastructure timeout or server error on our side, the execution credit is automatically refunded.
What happens if I exceed my monthly limits?
You can purchase addon credit packs or upgrade your tier directly from your billing panel to keep schedules running.
Is payment secure?
Yes, all payments are securely processed by Stripe or Dodo Payments, ensuring PCI-DSS compliant credit card handling.
Do you support volume enterprise plans?
Yes, we offer tailored SLAs, private VPC deployments, and custom pricing plans for large engineering teams.
What is your uptime guarantee?
We target a 99.1% uptime for our orchestration API. Our sandbox runners employ auto-scaling cluster groups to handle traffic spikes.
How do you handle API rate limits?
Remii includes built-in rate-limit queues. If a connected app responds with a 429 status code, Remii will pause and back off automatically.
Are watchdog services active on free plans?
Yes. All subscription levels benefit from our watchdog state machine recovery systems.
What happens during a global outage?
Our servers employ geo-redundant hosting. If a region fails, sessions failover to active nodes in secondary regions.
How do I check system status?
You can visit our live status dashboard at status.remii.space to monitor API, database, and sandbox runner health.
What is the timeout limit for a single script run?
Standard scripts have a maximum hard limit timeout of 120 seconds inside the container sandbox to prevent infinite loops.
Can I inspect historical job execution graphs?
Yes. The web console shows execution timelines, success rates, resource usage metrics, and error rates.
How are queue tasks prioritized?
Tasks are processed FIFO. Premium subscribers have access to dedicated higher-priority worker queues for instant starts.
Can I throttle execution speeds?
Yes. You can instruct your agent: 'Execute slowly with a 5-second pause between each tool action'.
How does Remii connect to the internet?
Sandbox containers have outbound internet access via secure NAT gateways with rotating IPs to facilitate scraping and API requests.
How do I get support?
You can email our customer helpdesk at support@remii.space. We offer chat support directly on our web dashboard for premium tiers.
Can I hire a consultant to set up Remii?
Yes. We offer enterprise setup packages where our engineers assist you in configuring custom internal toolkits and memory guides.
How do I report a bug?
You can open an issue on our GitHub repository or contact our team via support@remii.space.
How often is Remii updated?
We deploy updates, optimization patches, and new OAuth connectors weekly. You can track releases in our Changelog.
Can I request new integrations?
Absolutely. Vote on or suggest new connectors in our community integrations board.
Do you have tutorial videos?
Yes, check our documentation page and YouTube channel for walkthroughs on setting up Telegram triggers and databases.
Is there an onboarding call?
Enterprise and Team plans include a dedicated 30-minute onboarding call with our customer success team.
Where can I read user reviews?
Read case studies, user reviews, and developer testimonials on our website and community forums.
Supported Payments
We accept major credit cards globally via Stripe, and support localized direct payment methods like UPI for developers in India.