WrighteryContext
What your analysts know
Documents, data, knowledge, memory and integrations
Vibe working for analysis
You define the request.
Your AI analyst team builds the deliverable.
Vibe coding gave developers teams of coding agents. Wrightery brings the same model to knowledge work: your AI analysts research, cross-check and draft from trusted sources — you review the evidence and make the final call.
wrightery.com / workspaces / lumirit / runs
Grounded in the sources you trust
Your documents & filesDatabasesAPIsWeb researchMCP serversFile storage
How it works
Describe the work once. Run it every cycle.
Three steps, in plain language. No code, no node wiring, no waiting on engineering.
- 01You
Describe the work
Say what you produce and how you judge it — the way you'd brief a new analyst on their first day. Then point it at your documents, data and sources.
- Plain language
- Your sources
- Your standards
- 02Wrightery
Assemble the team
Wrightery builds the workflow and staffs it — specialist AI analysts with their own roles, sources and tools, plus the people who review and approve. Read it, change it; no engineer required.
- Specialist roles
- Explicit handoffs
- Yours to edit
- 03Your AI analysts
Research, review, decide
The analysts gather the evidence, cross-check each other and draft the deliverable, every claim traceable to its source. It stops at you before anything ships — and runs again next cycle.
- Every claim traceable
- Cross-checked
- You approve
One request, one team
Ask once. A team of AI analysts does the work.
Each analyst has one job and its own sources. They check each other before anything reaches you — and the run stops at your approval.
›Weekly memo: has anything changed in the thesis on our largest holding?Example
- Done
Research Analyst AI
Pulled the 10-Q, the earnings call and three supplier filings
- Done
Research Verifier AI
Checked every figure against its source — one flagged for review
- Done
Bear Case Analyst AI
Challenged the margin assumption the thesis rests on
- Waiting for you
You Reviewer
Review the evidence, then approve, send back or edit
Deliverable
Weekly memo.docx
Draft · 6 pages
- Every claim linked to its source
- Bull and bear cases side by side
- Held until you approve
ApproveSend back
An illustrative run. The steps are what the product does; the specifics are an example.
The workbench
Eight building blocks. One place to run the work.
Everything the work needs — what your analysts can read, what they can do, who reviews it, how it runs and what comes out — in one governed system instead of six tools you hold together yourself.
Capabilities
What they can do
Tools, APIs, queries and skills
Agents
Who does the work
A model, its context and its capabilities, with one role
Humans
Who steers and signs off
Approvals, reviews and feedback
Execution
How the work runs
Schedules and loops, every step recorded and costed
Artifacts
What comes out
Documents, articles, pages, apps and APIs
Workspaces
Where it comes together
People and AI analysts on one mission
The finished work, not a chat reply
Finished memos, reports and briefings, ready for your review.
A chatbot hands you a fragment. Wrightery produces the deliverable itself — a document, an article, a page or a dashboard — edited in place, with the evidence trail attached.
wrightery.com / workspaces / lumirit
AI analysis you can verify and control.
Get analysis backed by your sources, checked from multiple perspectives, and ready for your review. You decide what the AI can do and what needs your approval.
Built on your sources
Uses your approved documents, data, and tools to produce analysis relevant to your work.
Clear links to evidence
Trace claims back to their sources so you can check the evidence behind each conclusion.
Checked from multiple perspectives
Specialist agents challenge assumptions and flag conflicting evidence before you review the draft.
AI and humans in one workflow
Assign tasks to AI agents and team members, share feedback, and pass work between them—from research to final approval.
Clear limits on every task
Set budgets, permissions, and responsibilities. Track each step and its cost.
Your approval where it matters
Review and approve consequential actions before they go ahead.
You choose the level of autonomy
Let the AI recommend, act with approval, or work independently within the limits you set.
Who it's for
Analyst is a mode, not a job title.
Every knowledge worker is an analyst some of the time. If you produce the same kind of deliverable on a cadence, that is the work Wrightery is built to run.
Investment research
The weekly memo
Analysts track filings, news, fundamentals and sentiment, argue the bull and bear cases, and hand you one view with the reasoning trail attached.
Compliance & risk
The recurring review
The same review against this cycle's documents and records, every finding pointing at its source — and the exceptions escalated to a person.
Strategy & advisory
The client deliverable
A scattered source set turned into the briefing you already know how to write — to your firm's standard, in your firm's format.
Market intelligence
The landscape update
Competitors, pricing and coverage watched continuously; the analysis refreshed when something moves, not when someone remembers to look.
Your next report, with a team behind it.
Wrightery is in private build. Join the waitlist and we'll bring you in as soon as a seat opens — then describe one deliverable you already produce, and watch your AI analysts research it, cross-check it and draft it for your review.
No credit card. We'll email you the moment your seat is ready.
Rather talk it through? Book a 30-minute meeting ↗