Turn 3 hours of email into 20 minutes.
ReplyrX drafts every reply inside Outlook using your business knowledge and your voice — so you review, not rewrite.
You're spending hours writing replies
that should take minutes.
2–3 hours lost to email every day
Writing thoughtful, accurate replies takes time you don't have. You're constantly switching from strategic work to drafting, and back.
Generic AI doesn't know your business
You paste the email into ChatGPT, get a decent draft, then spend another 5 minutes adding your actual pricing, policies, and context. Every. Single. Time.
Your team writes differently — and inconsistently
New hires don't know the right tone. Sales replies don't match support replies. Each email is a roll of the dice for how your company actually sounds.
ReplyrX knows your business.
Generic AI doesn't.
It lives inside Outlook's sidebar. It reads the email you're looking at. Then it drafts a reply using your actual pricing, policies, and tone — not a hallucinated guess.
Open any email in Outlook
ReplyrX reads the full context automatically — subject, sender, thread history.
Pick your reply style
Choose from 9 built-in profiles: Sales, Executive, Firm, Friendly, and more. Or create your own.
Get a draft. Refine it. Insert.
The AI drafts using your business knowledge. Tweak it conversationally, then insert with one click.
From inbox to polished reply
in under 30 seconds.
Watch ReplyrX read a real customer email, pull from the Knowledge Base, and generate a complete reply — live.
Everything you need to reply faster,
without leaving Outlook.
The AI that actually knows your business facts.
Store your pricing, policies, contacts, and deadlines once. ReplyrX automatically injects the relevant facts into every draft — you never have to repeat yourself again.
Replies in your voice, not a generic one
Choose from 9 built-in profiles — Sales, Executive, Firm, Friendly, Technical, and more. Each profile is a fully configurable AI persona with its own instructions and model.
Refine conversationally until it's right
"Make it shorter." "Add the deadline." "Change to French." Multi-turn refinement means one draft is usually all you need — with a quick tweak or two.
Not locked into one AI provider
Use GPT-4.1 for executive replies, Claude for nuanced client situations, and a faster model for quick acknowledgements. Each profile runs on the model you choose.
Dictate your instructions by voice
Don't want to type? Speak your reply instructions. Whisper-powered transcription captures exactly what you said, live, before drafting your reply.
Up and running in three steps.
Install the add-in
Install via Microsoft AppSource or your M365 Admin Center. Works on Classic Outlook, New Outlook, Outlook Web, and Mac.
Add your business knowledge
Import your pricing, policies, and facts from a spreadsheet — or add them one by one. Team members can suggest facts for admin review.
Draft, refine, send
Open any email, pick a profile, and generate a draft. Refine it with plain instructions. Insert into your reply with one click.
Everything you need to know
before your first draft.
Quick answers to the most common questions. For the full walkthrough, read the User Guide →
From the admin portal dashboard, scroll to "Get the Outlook Plugin." You can install it yourself by downloading the manifest.xml and sideloading it via Outlook → Get Add-ins → My Add-ins → Add from file. Or, your IT admin can deploy it to the whole team via the Microsoft 365 Admin Center.
The Knowledge Base is the set of facts the AI consults when drafting — your pricing, policies, product details, FAQ answers. The richer it is, the better every draft will be. Any team member can suggest facts; admins approve them before they go live.
A Profile controls how the AI writes — tone, style, model, and behaviour rules. A Template is static text you insert and customise. Use profiles for AI-generated replies; use templates when the wording must be exact (legal, compliance, scheduling).
No. ReplyrX never stores email content. The Usage Log records only metadata — who generated a draft, which profile was used, which model, and when. The portal header explicitly notes "metadata only, no email content stored."
Three things make the biggest difference: (1) fill the Knowledge Base with accurate facts, (2) use the Additional Instructions field to give the AI specific context for each email, and (3) match the profile to your audience — not just the topic.
ReplyrX supports OpenAI (GPT-4.1), Anthropic (Claude), Google (Gemini), and xAI (Grok). Each profile can be configured to use a different model and provider — so you can use a faster model for quick replies and a more capable one for complex negotiations.
Only Admins can approve, reject, bulk-approve, or edit approved entries. Any team member (Member or Admin) can suggest facts and bulk-import entries — but they land in Pending Review until an admin approves them.
Yes — click the pin icon at the top right of the plugin panel. Once pinned, ReplyrX stays visible as you move from one email to the next, which is the recommended setup if you reply to many messages in a row.
Be first when
ReplyrX goes live.
We're bringing ReplyrX to Microsoft AppSource this week. Join the waitlist and we'll notify you the moment it's ready to install.
Join the Waitlist