Skip to content
AI writing studio · for LinkedIn

Sound like
yourself. Finally.

Verso writes LinkedIn posts from your real work: yourdocs, your notes, your past posts. So it reads like you wrote it. And it never invents a number it can't source.

10 posts/mo free No credit card Encrypted by defaultPublishes via LinkedIn's official API
Verso Studio
GROUNDED
tone · candidformat · storyhook · contrarian
MC
Maya Chen
Founder, Loomwork · now ·
We shipped 3 things in Q3 nobody asked for. Two flopped. One, a 4-line onboarding change, quietly lifted activation 22%. The lesson wasn't "ship more." It was that our roadmap was guessing and our users weren't. So we started reading the retros, not the roadmap. Grounded beats clever. Every quarter.
2 sources
this post50 words
grounded in Q3-retro.pdf · founder-notes.md

Built like infrastructure, not a wrapper.

Official LinkedIn API
Powered by Claude · Anthropic
AES-256-GCM at rest
Tenant isolation enforced in the database
Your content never trains models
Grounded in your real documents
Flat plans · no surprise bills
Grounded generation

Drafts cite your real numbers, not invented ones.

Drop in PDFs, DOCX, slides, notes, CSVs, even screenshots, which are read on the way in. Everything is indexed into your private, encrypted knowledge base. Every draft retrieves the most relevant passages from your own material, so the stats and stories trace back to what you actually did.

  • If it isn't in your knowledge base, it doesn't go in the post: the no-fabricated-metrics rule.
  • Passages are matched by meaning, not keywords, so the right story surfaces.
  • Screenshots and images are read with vision OCR on the way in.
studio · grounding on
GROUNDED
generic · ungrounded

“We boosted engagement by an impressive amount last quarter.”

grounded · from your work

“Cut onboarding time 40% in Q2 - down from 11 days to 6.”

retrieved from: Q2-retro.pdf · chunk 4/12
How it works

Three steps from your work to a post in the feed.

01

Upload your work

Drop PDFs, DOCX, notes, CSVs, even screenshots. Claude vision OCR reads them into your private, encrypted knowledge base.

02

Dial in the controls

Pick tone, hook, format, and length. Set emoji and hashtag levels, add a CTA, toggle knowledge-base grounding and live web research.

03

Generate grounded posts

Watch the draft stream into a live LinkedIn preview, then save, copy, regenerate, rewrite, expand, or shorten in one click.

The command center

Control every dial. Watch it land before you post.

Five tones, five hook styles, four formats, three length bands, plus emoji and hashtag levels, an optional CTA, knowledge-base grounding, and live web research. As the draft streams, a pixel-faithful LinkedIn preview assembles on the right.

  • Desktop and mobile preview, rendered exactly as the feed shows it.
  • Real “…see more” truncation, reaction dots, and a live streaming cursor.
  • Save, copy, regenerate, rewrite, expand, or shorten in one click.
/studio
READY
Idea

What I learned shipping our RAG pipeline in 6 weeks

Tone

ProfessionalBoldStorytellingContrarianInspirational

Hook

QuestionBold claimStatisticStory openContrarian

Format

Short takeStoryListicleHow-to

Length

ShortMediumLong
KB grounding Web research Emoji: minimal Hashtags: few
Generate post
Topic research

Hand it a topic. Get back a sourced, on-brand draft.

It runs live web research with built-in search and blends what it finds with your knowledge base. The post comes out both current and unmistakably you: trend-driven content without the blank-page tab-hopping.

/research
LIVE

Topic

The state of RAG in production, 2026

Live web search · 5 sources · blended with your knowledge base
arxiv.org engineering.blog docs.anthropic.com thenewstack.io

Draft · on-brand, sourced

“Everyone says RAG is solved. In production, three things still break first - and we hit all three shipping ours…”

Fair usage

No surprise bills.
Ever.

Most AI tools hand you a mystery bill on the 1st. Verso is a flat monthly plan with clear included usage: meters you can read at a glance, a warning before you ever reach a limit, and a price that never moves mid-month.

1 flat price
per month, per plan
0
bill surprises
See your plan meters
Plan usage · Freelive
Posts & chat6 / 10 this month
Research drafts1 / 3 this month
Monthly allowance48% used
everything resets on the 1stheads-up at 80%, never a lockout without warning
Privacy & ownership

Your knowledge base is encrypted before it ever reaches the database.

Upload your CV, client docs, and past posts. Each piece is sealed with AES-256-GCM on the server (using a key that lives only in your environment) before a single row is written.

Own your private AI LinkedIn studio: data, keys, and bill all yours.

Encrypted at rest

AES-256-GCM, authenticated. Every sensitive field is sealed server-side before a single row is written, and the encryption key is held apart from the database. The database only ever stores ciphertext.

Tenant isolation, enforced by Postgres

Row-Level Security is FORCED on every tenant table and keyed to your org per transaction. A forgotten WHERE clause physically cannot read or write another tenant's data.

Audited & governed

An append-only audit log records sensitive actions with originating IP. Roles and memberships, per-plan usage limits, and rate limiting are built in.

Never trains a model

Your work grounds your drafts and nothing else. No training, no sharing, no third-party resale. It's your private studio, and everything in it stays yours.

drizzle/0001_rls.sql
ENFORCED
-- forced even for the table owner
ALTER TABLE "kb_chunks" FORCE ROW LEVEL SECURITY;
CREATE POLICY "kb_chunks_tenant_isolation" ON "kb_chunks"
  USING      (organization_id = current_setting('app.current_org_id', true))
  WITH CHECK (organization_id = current_setting('app.current_org_id', true));

Isolation is enforced by Postgres, not by hoping every developer remembers.

Straight promises, before you sign up.

Verso only writes from what you give it - your documents, notes, and past posts - so a draft sounds like you and points back to work you actually did. That is the whole product.
The core promise
Flat plans and plain usage meters. You always know where you stand, we warn you before a limit, and the price never moves mid-month. No surprise bills.
On pricing
Your uploads are encrypted at rest, never sold, and never used to train models. It is your private studio, and everything in it stays yours.
On privacy
Pricing

Start free. Upgrade when paid plans land.

The free tier is live today. Pro and Team are coming soon. The studio, knowledge base, and live cost dashboard are all yours in the meantime.

Current planFREE
£0forever
10 posts a month grounded in your real work - the full product, sized to try.
No payment method on file
This month's usage
Generations0 / 10
Research drafts0 / 3
Plan allowanceunlimited
Everything your plan includes, with no overage charges - ever.
Free · current
£0
10 posts a monthFull grounding in your workUsage meters, no card needed
Start free
Starter
£9/mo
30 posts a month6 carousels a monthResearch drafts included
Pro
£29/mo
Unlimited posts + carousels50 images · 15 videosPriority queue
Team
£99/mo
Unlimited everythingShared workspace + rolesBudget controls + SSO

Questions, answered honestly.

Your next post is four minutes away

Write the post only you
could've written. Verbatim.

10 posts a month, free · no card · upgrade only when you outgrow it