Eloquens AI
Eloquens AI is an autonomous email agent that plugs into a shared business inbox, reading, drafting, sending and routing messages around the clock. Replies are grounded in your own knowledge base and written in 160+ languages.
What is Eloquens AI?
Eloquens® AI is an autonomous email agent that plugs into a real, working mailbox and processes it end to end. The product rests on one central idea, the Email Agent: an AI identity connected to an inbox that reads every incoming message, decides whether it can respond, drafts a reply grounded in your knowledge base, and either sends it automatically or places it in your draft queue for review. Each email is handled in five minutes or less, 24/7. The agent reads and answers in the sender's own language across more than 160 languages, corrects grammar, and mirrors the sender's tone and style; a custom voice can be tuned per agent to fit a given use case. The vendor claims thousands of email conversations are handled simultaneously across all departments. Answers are not improvised. Each one is grounded in a knowledge base you supply: uploaded documents, FAQs, product sheets and policies, or connected sources such as Google Drive and OneDrive, with no storage cap on any plan. Eloquens states the agent searches your knowledge base to ground its reply in accurate, approved content, so it never makes things up. When a question falls outside what that content can confidently cover, the agent escalates under your rules, flags a partial draft for review, or routes sensitive subjects straight to a human without attempting an answer. It also analyzes sender sentiment and intent, books appointments, filters spam and newsletters, and logs the reasoning behind every message it sends in a full audit trail. You decide the automation level: full autonomy once the agent is trained, or drafts held back for human approval. Mailboxes connect through Gmail (Google Workspace), Microsoft 365/Outlook and Kerio Connect; other platforms go through the Enterprise offer. The privacy policy names OpenAI and Anthropic as the third-party AI engines behind the scenes. Billing is unusual and favorable: only sent emails count against your quota, while reading, drafting and routing decisions are free. The publisher is Ignite Enterprise Software Solutions, LLC, trading as IgniteTech, an ESW Capital group company based in Austin, Texas, which announced the product through a PR Newswire release. It is recent: the domain was registered in September 2024 and first archived in January 2025. It runs as a web app on app.eloquens.ai, with no mobile app, and you can test it live by emailing the AI persona ella.quinn@eloquens.ai.
What it does
- Reads and understands every incoming email
- Drafts a reply grounded in your own knowledge base
- Sends automatically or queues a draft for human review
- Routes and escalates messages to the right person under your rules
- Replies in the sender's own language, across more than 160 languages
- Filters out spam, newsletters, marketing campaigns, auto-replies and system notifications
- Records the reasoning behind every sent email in a full audit trail
When to use Eloquens AI / When not to
A quick filter to help you decide if Eloquens AI is the right fit.
When to use Eloquens AI
- Support, customer service, HR and administrative teams handling a high volume of repetitive email on shared inboxes
- Multilingual organizations that need every sender answered in their own language, across more than 160 languages
- Small and mid-market companies sized for the self-service plans: one agent on Solo, up to three on Team, up to ten on Business
- Independent professionals and freelancers running their own inbox on the entry-level Solo plan
- Contact centers and large accounts that need dedicated infrastructure, custom AI training and a contractual SLA through the Managed Deployment offer
When not to use Eloquens AI
- Organizations processing health data: the terms of use expressly forbid using the service for Protected Health Information as defined by the HIPAA Privacy Rule, whatever the healthcare use-case page suggests
- Workflows that collect sensitive identifiers, which the acceptable use policy prohibits: driving licenses, passports, social security or tax numbers, banking details and health information
- Fully automated decision-making with no review mechanism, and personalized legal, medical or financial advice issued without a qualified professional in the loop, both barred by the acceptable use policy
- Anyone looking for a free consumer tool: there is no permanent free plan, only a 30-day trial, and inboxes outside Gmail/Google Workspace, Microsoft 365/Outlook and Kerio Connect require the Enterprise offer
- Organizations that require European data residency: hosting and transfers are in the United States, with no EU option on offer
How to use Eloquens AI
A typical end-to-end flow, from setup to results.
- Test the agent without registering by emailing ella.quinn@eloquens.ai, an AI persona that answers you directly
- Register on app.eloquens.ai to start the 30-day trial on the Solo plan, with no credit card required
- Connect your existing mailbox: Gmail (Google Workspace), Microsoft 365/Outlook or Kerio Connect, which typically takes minutes with no IT involvement required
- Set up your virtual email assistant the way you would onboard a new team member
- Define your AI personas and how each type of email should be handled, from a simple reply to a complex workflow
- Load the knowledge base: upload documents, FAQs, product sheets and policies, or connect Google Drive and OneDrive
- Write your filtering, routing and escalation rules to match your own criteria
- Choose the automation level: full automation, or drafts generated and held for human review
- Manage live conversations, review drafts and take back control of any exchange at any time
- Monitor performance on the analytics dashboard: volume, resolution rate and response time
Pros & Cons
Pros
- A favorable billing unit: only sent emails count, while reading, drafting and routing are free, and spam, newsletters, marketing campaigns, auto-replies, out-of-office messages, system alerts, social notifications, order confirmations, payment receipts and delivery failure notices are filtered out and never counted
- 30-day trial on the Solo plan with no credit card required, and no setup fee on any plan
- Unlimited knowledge base with no storage cap, and 160+ languages at no extra cost, on every plan
- Full audit trail documenting the reasoning behind each sent email, which is uncommon in this segment
- You control the automation level: drafts for review, or fully autonomous sending once the agent is trained
- No surprise overage: alerts at 80% and 100% of quota, sending simply stops, and 99 USD email packs never expire
- An established, identifiable publisher, IgniteTech of the ESW Capital group, with a public address and DPO, EU/UK/Swiss Data Privacy Framework certification, and an explicit commitment that Google Data is never used to train AI models
Cons
- Two contradictory price lists sit on the same homepage, a Pricing 2026 block (Solo 60, Team 150, Business 300 USD per month) and a legacy block (Professional 65, Advanced 169, Business 425 USD per month), leaving the price actually charged uncertain
- No permanent free plan, and no refunds: payments are non-refundable, including annual plans paid up front
- Automatic renewal with 60 days' termination notice, and a renewal price disclosed only on written request 90 days before the term ends
- Section 5.2 of the terms allows de-identified aggregate data to be used to train AI models, with no documented way for the customer to opt out
- Self-service email connectors limited to Gmail/Google Workspace, Microsoft 365/Outlook and Kerio Connect, and no mobile app
- An API usage right is granted in the terms but no public API documentation exists, and there is no contact page, no published support email and no public subprocessor list
- United States hosting and transfers with no European residency option, SOC 2 and ISO 27001 claimed for the hosting infrastructure without any published document or trust page, and Texas law with exclusive jurisdiction in Austin
Pricing & Plans
There is no permanent free plan. Eloquens AI is sold on subscription, priced in US dollars and exclusive of applicable taxes, with a 30-day trial on the entry-level plan that requires no credit card. The lowest advertised entry point is 60.00 USD per month for the Solo plan, under the Pricing 2026 list. It should be noted that a legacy price list is still displayed on the same homepage, starting at 65 USD per month for the Professional plan, or 58.34 USD per month billed annually, so the applicable rate should be confirmed in writing before subscribing. Annual billing is invoiced as ten months paid up front, giving two months free, with the discount applied automatically. Only sent emails are metered, whether a direct reply or a forward with generated context; reading, drafting and routing decisions are free. Additional volume is sold as a one-off 99 USD Email Pack covering 1,000 extra sent emails, with no expiry and no overage billing. There are no setup fees on any plan. Monthly subscriptions may be cancelled at any time with access until the end of the current period, whereas annual plans are paid in advance and are non-refundable. Upgrades apply immediately on a pro-rata basis and downgrades take effect at the next period. Managed Deployment and partner-channel arrangements are quoted on request.
- the Pricing 2026 list meters sent emails
- while a legacy list still on display meters incoming emails and caps knowledge storage. Confirm which one applies before subscribing.
- 60 USD per month
- 1 Email Agent
- 500 sent emails per month
- for individual professionals and freelancers managing their own inbox
- with a 30-day trial
- 150 USD per month
- up to 3 Email Agents
- 2
- 000 sent emails per month
- for growing teams running multiple inboxes with shared oversight
- flagged Most Popular
- 300 USD per month
- up to 10 Email Agents
- 5
- 000 sent emails per month
- for mid-market companies running email at scale across departments
- adding white labeling
- priority support with SLA
- onboarding and setup
- 160+ languages
- routing rules
- sentiment analysis
- autonomous replies
- full audit trail
- integrations
- analytics dashboard
- unlimited knowledge base and custom voice
- 99 USD one-off purchase for 1
- 000 additional sent emails
- with no expiry
- priced on quotation
- aimed at contact centers
- large enterprises and partners
- 65 USD per month (58.34 billed annually)
- 1 user
- unlimited workflows
- 1 workspace
- 1
- 000 incoming emails
- 250 MB of knowledge storage
- 169 USD per month (150 billed annually)
- unlimited users
- 5 workspaces
- 3
- 000 incoming emails
- 500 MB
- conversation translation and an onboarding session with a specialist
- 425 USD per month (383.34 billed annually)
- unlimited users
- 10 workspaces
- 8
- 000 incoming emails
- 1 GB
- conversation translation and an onboarding session
- GFI Software customers are directed to their reseller to secure a license
Data, GDPR & hosting
A consolidated view of how Eloquens AI handles your data.
GDPR overview
GDPR compliance support is advertised on the homepage and backed by concrete arrangements. The privacy policy dated 29 July 2026, hosted on ignitetech.ai and linked from the Eloquens footer, states that the Ignite U.S. Corporate Group, which includes Ignite Enterprise Software Solutions, LLC, is certified under the EU-U.S. Data Privacy Framework, its UK extension and the Swiss-U.S. DPF, under Federal Trade Commission jurisdiction. Legal bases cited for the EEA, UK and Switzerland are consent, performance of the contract and legitimate interests, with prior consent for sensitive data. Rights offered include access, rectification, erasure, transparency, objection, restriction, portability and withdrawal of consent, with a 15-day appeal window and recourse to the relevant supervisory authority, the ICO or the Swiss FDPIC; DPF disputes go to BBB National Programs free of charge. A data protection officer is reachable at privacy@ignitetech.com. No EU representative under Article 27 is named.
Who owns the data?
Section 2.1 of the terms of use is unambiguous: the customer retains all ownership rights in the Input it submits and owns all Output, with the service provider assigning any rights it might hold in that Output. Customer Inputs are treated as the customer's confidential information under section 6.1. The provider keeps ownership of the software and its materials (section 5.1) and owns the de-identified Aggregate Data, statistics, benchmarking and analytics derived from performance and usage (section 5.2); any suggestion or improvement the customer submits becomes the provider's property (section 6.6). For data processed on behalf of a business customer, the privacy policy makes that customer the controller and IgniteTech the processor.
Reuse rights
Two clauses have to be read together. Section 2.2 restricts the provider to using Customer Content only as necessary to deliver the service, comply with applicable law and enforce its own policies. Section 5.2 then states that Aggregate Data, de-identified so the customer cannot be identified as its source, may be used to train artificial intelligence models. Customer content is therefore excluded from training while aggregated usage data is not, and no customer-controlled opt-out is documented anywhere on the site. Requests are handled by two named third-party AI providers, OpenAI and Anthropic, under commercial API terms that exclude customer data from foundation-model training. Google Data from Gmail, Drive, Calendar and Contacts is used solely to run the service, is not read by humans except for security, legal or consented reasons, is never used for advertising, and is expressly excluded, derivatives included, from developing, training, fine-tuning or improving any AI model. As the customer owns all Output under section 2.1, it can reuse and exploit generated replies freely, without asking permission. Site visitors are also subject to automatic collection of IP address, cookie identifiers, browser, system, pages visited and, where applicable, geolocation, with Google Analytics in use.
Data retention & training
Hosting summary
Eloquens AI does not name a single data center or region as the storage location for mailbox content. The homepage only states that the hosting infrastructure is SOC 2 and ISO 27001 compliant, without publishing a certificate or a trust page. What is documented is the jurisdiction: the Ignite U.S. Corporate Group receives and processes personal data in the United States, under Federal Trade Commission oversight. Personal data originating from the EEA, the United Kingdom and Switzerland is therefore transferred to the United States, with the EU-U.S. Data Privacy Framework, its UK extension and the Swiss-U.S. DPF serving as the transfer mechanism. No European data residency option is offered. Two third-party AI providers are named as processing user requests: OpenAI and Anthropic. The full subprocessor list is not published and is supplied only on request. The public web infrastructure sits behind Cloudflare, whose anycast network answers from an Austin node, which tells you nothing about where customer content itself resides.
Where Eloquens AI works
Country-level availability.
Not available in
Things to keep in mind
Risks and trade-offs to weigh before adopting Eloquens AI.
- Two contradictory price lists on the same homepage: Pricing 2026 (Solo 60 / Team 150 / Business 300 USD per month, quotas counted in sent emails) and a legacy list (Professional 65 / Advanced 169 / Business 425 USD per month, quotas counted in incoming emails, with capped knowledge storage). Confirm the applicable rate in writing before subscribing.
- A HIPAA contradiction to take seriously: the sector use-case page promises healthcare customers HIPAA-compliant confidentiality, while section 4.5 of the terms of use expressly forbids creating, receiving, maintaining, transmitting or otherwise processing Protected Health Information through the service.
- A mixed position on model training: section 2.2 limits Customer Content to service delivery, legal compliance and policy enforcement, but section 5.2 allows de-identified aggregate data drawn from your usage to be used to train artificial intelligence models, with no customer-controlled opt-out documented.
- Renewal traps: automatic renewal, 60 days' notice to terminate, a renewal price disclosed only on written request 90 days ahead, non-refundable payments including prepaid annual plans, and an all-caps clause warning that late notice, or notice sent to the wrong provider email address, does not terminate the agreement.
- Obligations that are easy to overlook: a publicity clause committing you to cooperate on a press release and a testimonial advertisement and to display a Powered by logo on pages using the software, plus an annual license-compliance audit whose costs are recharged to you beyond 5% unlicensed usage, all under Texas law with exclusive jurisdiction in Austin.
- Trust claims you cannot verify: SOC 2 and ISO 27001 are asserted for the hosting infrastructure with no trust page or certificate published, no public subprocessor list (supplied on request only), no EU Article 27 representative named, and a generic IgniteTech group privacy policy covering all products rather than Eloquens AI specifically.
- The human risk of handing over the inbox: the terms disclaim any warranty as to the accuracy of Outputs, so unreviewed autonomous replies can carry mistakes sent in your name and bind your relationships; teams that stop writing and arbitrating lose the very judgment the agent relies on when it escalates; and the acceptable use policy requires you not to conceal from correspondents that they are talking to an AI when that is not obvious.
Setup & Integrations
Technical difficulty
Low. No development skills are required: connecting Gmail/Google Workspace, Microsoft 365/Outlook or Kerio Connect runs through OAuth consent and, per the vendor, typically takes minutes with no IT involvement required. The real effort is configuration rather than technology: defining personas and workflows, writing routing rules, and loading the knowledge base from uploads or connected Google Drive and OneDrive. Autonomous replies only switch on once the agent is trained, so budget a supervised run-in period. The Business plan includes onboarding and setup, and Managed Deployment hands the whole build to the vendor. An API exists but has no public documentation.
Deployment
Integrations
Behind Eloquens AI
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is an Email Agent?
What counts against my monthly quota?
What happens when the AI cannot answer?
Which email providers are supported?
What goes into the knowledge base?
What happens when I reach my quota?
Is there a setup fee, and can I cancel?
How does the free trial work?
Is my data used to train AI models?
Where is my data hosted, and is a DPA available?
Should you pick Eloquens AI?
Eloquens AI tackles one clear problem: the repetitive inbound email that eats a team's day. Its answer is an autonomous agent grounded in your own knowledge base, with an optional human safety net through draft mode, so nothing has to leave the building unreviewed. Three things genuinely set it apart: only the emails actually sent are billed, the knowledge base has no storage cap on any plan, and a full audit trail records the reasoning behind every reply. For a support, service or HR team facing a high volume of standardized questions, that combination is convincing. Check three things before signing. The price first: two contradictory lists sit on the same homepage, one metering sent emails from 60 USD per month, the other metering incoming emails from 65 USD, so ask in writing which one applies to you. The contract next: automatic renewal, 60 days' notice to terminate, no refunds even on prepaid annual plans, Texas law and Austin courts. Then the data: hosting and transfers are in the United States with no European residency option, the subprocessor list is available only on request, and while Customer Content is limited to service delivery, de-identified aggregate data may be used to train AI models. Anyone in healthcare should note that the terms of use forbid processing protected health information, whatever the use-case page suggests. The product is young, with a domain registered in September 2024, but the publisher is not: IgniteTech, an ESW Capital company, is established and identifiable, with a named DPO and Data Privacy Framework certification. Eloquens AI makes sense for a team that handles a large volume of standardized email, wants to keep control of the automation level, and accepts American hosting. Use the 30-day trial before committing to a year.
- Choosing a selection results in a full page refresh.
- Opens in a new window.