Roadmap
Last updated on April 22, 2026
What we're building next, grouped by how close it is to shipping.
Vellum is moving fast. This page shows where things stand. We update it as priorities shift, so check back.
How to read this
Four buckets, ordered by proximity:
- Shipping soon — engineering is mostly done, rollout is imminent.
- Up next — actively in progress, unblocks the next level of usefulness.
- On the horizon — scoped and prioritized, not yet started in earnest.
- Exploring — ideas we believe in but haven't committed dates to.
Shipping soon
Engineering-complete or nearly so. Finalizing for rollout.
🪄Teleport
Your assistant doesn't go dark when your laptop closes. Move seamlessly between local and cloud — same memory, same identity, same conversations.
🌐Cross-channel context
Start a conversation on desktop, continue it on Slack, finish it on your phone. One assistant, one memory, wherever you are.
🔧Setup, upgrade, recovery
First-run should feel like meeting someone, not configuring software. Smoother installs, safer upgrades, and a recovery path when something goes sideways.
✉️Email autonomy
Your assistant drafts, replies, and handles routine email on its own — with the guardrails you set. You stay in the loop; it does the busywork.
🎥Google Meet
Join meetings, take notes, track action items, and pull them into your to-dos — without another tab to manage.
Up next
In progress now. These are the things that unlock a meaningfully better day-to-day.
📱iOS app
A real mobile assistant, not a wrapper. Native iOS, push notifications that matter, and the same memory your desktop assistant has.
🌍Web app
Access your assistant from any browser, on any device, without an install. Same assistant, zero friction.
🤖Slack autopilot
Full Slack presence — your assistant watches the channels you care about, replies where you'd want it to, and escalates the rest.
🔒Permissions, rebuilt
A permissions model that's both safer and less annoying. Clear defaults, granular overrides, and no more guessing what you approved a week ago.
⏰Schedules without tokens
Recurring tasks and check-ins that don't break when an API token expires. Reliable background work, handled behind the scenes.
🎯Personalization
Your assistant adapts to how you work — tone, habits, preferences, the people you talk to. It gets more yours the longer you use it.
🧹Automatic context compaction
Long conversations stop hitting a wall. Your assistant keeps the signal and drops the noise, automatically.
🧠Subagents
Your assistant delegates to specialized helpers in parallel — research, coding, writing — and brings the results back to you.
📈Automatic self-improvement
Your assistant notices when it got something wrong and quietly gets better at it. No fine-tuning pipeline required.
🩺Vellum Doctor
A built-in diagnostic that tells you what's broken and how to fix it — before you have to ask.
💻Better CLI
Power-user shortcuts, scriptable commands, and a faster path for developers who want to drive Vellum from the terminal.
📬Native email accounts
Assistants can own their own email accounts, not just borrow yours. Real identity, real inbox, real separation of concerns.
⭐Vellum Pro plan
A paid tier for heavy users and small teams — higher limits, priority support, and features built for people who use Vellum every day.
🗺️Roadmap in docs
The page you're reading. Public, evolving, and honest about what we're building next.
On the horizon
Prioritized, scoped, and waiting their turn. Expect motion here in the coming quarters.
🎙️Real-time voice
Talk to your assistant out loud, interrupt it, and have it talk back — with low enough latency that it feels like a real conversation.
📞Phone calls
Your assistant can take and make phone calls on your behalf — screening, scheduling, and handling the calls you don't want to.
💚Health & wellness
Pull signals from the wearables you already have and turn them into useful nudges, not a dashboard you have to check.
🔍PR review
Your assistant reviews pull requests with your standards, catches regressions, and flags the things you'd catch yourself.
🏗️Orchestrating coding agents
One assistant that coordinates multiple coding agents across your repos — planning, delegating, and shipping end-to-end.
🔗Webhooks
Wire any service into your assistant with a simple webhook. If it can POST, your assistant can react to it.
🚀Public apps & shareable pages
Ship something your assistant built with a single share link — dashboards, tools, little apps — no deploy step required.
🤝Skill sharing
A community hub for teaching your assistant new tricks. Pull in skills others have built; publish your own.
👥Multiple assistants per user
A work assistant and a personal assistant. Or one per project. Separate memories, separate identities, same you.
📍Location services
Your assistant knows where you are (when you want it to) and acts accordingly — timezone-smart, traffic-aware, context-aware.
💳Agentic payments
Your assistant buys the flight, renews the domain, tops up the subscription — within limits you set and receipts you can audit.
💬Assistant-to-assistant
Your assistant can talk to other assistants — coordinating schedules, sharing context, negotiating on your behalf.
🐦Twitter & social
Draft, schedule, and publish to your socials — in your voice, with your edits, never without your sign-off.
📝SEO content workflow
From keyword to published post, your assistant runs the whole content loop — while you stay the editor-in-chief.
👀Team visibility
Shared context across a team, without sharing a single assistant. Your memory stays yours; the work gets coordinated.
Exploring
Ideas we're excited about but not ready to commit timing on. Directionally yes, specifically TBD.
💬iMessage
A native Apple-ecosystem channel for your assistant — texts, read receipts, the works.
💰Personal finance
Your money, summarized and acted on. Categorization, alerts, and small automations you'd actually run.
🧬Own weights
Models fine-tuned on you, with you. A long-term path toward an assistant that's truly yours at the model layer.
⌚Apple Watch
Your assistant on your wrist — glanceable, taptic, and ambient.
📡Presence detection
Your assistant knows when you're at your desk, in a meeting, or asleep — and behaves accordingly.
🎤Voice activation
Wake-word access so your assistant is one phrase away, not one click away.
🖥️Ambient context
Your assistant can see what's on your screen when you ask it to — so "this" and "that" actually mean something.
⚡Use Vellum in any app
Summon your assistant inside any app on your system — summarize a doc, reply to a message, translate a selection — without switching windows.
🧑💻Open-source coding models
First-class support for running coding tasks on open-source models you control.
✨Dynamic visual presence
An assistant that can show up where it's useful — a floating panel, an inline overlay, a quiet notification — and get out of the way when it isn't.
🗄️Custom app backends
Let the apps your assistant builds talk to real backends you own, securely.
🏠Personal life automation
The small recurring stuff — birthdays, renewals, reservations, errands — handled without a spreadsheet.
🎨Custom assistant skins
Make your assistant look and feel the way you want it to. Themes, avatars, and the small cosmetic details that make something feel like yours.
📦Local containerization
A stronger security boundary between your assistant and the rest of your system, for users who want extra peace of mind.
☎️Managed phone numbers
Your assistant gets its own phone number, managed by us, ready to call and text on your behalf.
The fine print
Polish, bug fixes, and security improvements are a standing commitment — a meaningful slice of every team's capacity goes here, always. That work doesn't show up on this page, but it ships continuously.
Priorities shift. Dates move. Some items will merge, some will split, some will get dropped. This page changes with us. If there's something you want to see sooner, tell us — we read every note.
Want to share feedback on our Roadmap? Drop us a note in Discord.