Opportunities › AI research agents for Wall Street financial analysis › Proposed app
Deskbook
AI research workspaces that keep your findings, data and context between sessions
- Score
- 33
- confidence 45
- Evidence
- 7
- items on the page
- Rivals
- 3
- named in sources
- First version
- 45 d
- estimate
- Type
- Desktop app
- PWA: runs in the browser, installs on Mac and Windows
- Market since
- 2025
- New market · AI estimate
ScreensConcept
Desktop app: Investment analysts work through long research sessions managing portfolios, reviewing findings, and updating contexts—extended desk-based analytical work. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
Acme Corp coverage
Context loadedDay 14
About this appAI proposal
Deskbook gives investment analysts a research workspace that an AI agent can work in over days or weeks. Each workspace holds your portfolio context, data sources, findings and memory files, so you stop re-pasting context every time you return to a name.
Instead of loading long tool schemas, Deskbook generates small typed tool modules for the data sources you connect. Large price histories stay in the workspace as files. The agent reads summaries and pulls only the slices it needs.
When earnings land or a competitor reports, open the workspace and ask for an update. The agent sees what you concluded before, what changed, and which notes are now stale. Built for analysts, portfolio managers and small research teams.
- Tagline
- Research that remembers where you left off.
- Voice
- PreciseCalmPlainspoken
- Why this name
- A deskbook is the reference an analyst keeps within reach. The name suggests a standing research desk that stays organized between sessions.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Persistent research workspaces
One workspace per company, theme or portfolio. Files, findings and memory carry over between sessions, so nothing needs re-pasting.
No member yet
Typed tool modules
Connected data sources become compact generated tool modules, not verbose schemas that use up the context before work starts.
No member yet
Data kept out of the context
Long price series and filings are saved as workspace files. The agent works from summaries and queries slices on demand.
No member yet
Portfolio context built in
Add holdings, benchmarks, theses and watchlists once. Every agent run starts from them.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#14532DLedger Green · primary#F7F5EFPaper · surface#1C1917Ink · text#B7791FBrass · accent#3B5B7ASlate Blue · data#E7E2D6Stone · border
Typography
Aa
Deskbook
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test. Data vendor fees and compute costs are not known and may change the final price.
Trial
Free
- 1 workspace
- Sample datasets
- Limited agent runs
- Export to PDF
Analyst
$79 a month
- 1 seat
- Unlimited workspaces
- Connect your own data keys
- Memory and findings log
- Share links
Team
$59 a month
- Per seat, 3 seat minimum
- Shared workspaces and roles
- Shared memory
- Usage reporting
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$190,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Analysts export research briefs and update notes as shareable links or PDFs for PMs, clients and co-investors. Each carries a small 'Prepared in Deskbook' mark and a view of the findings history.
- Do they become customers?
- Recipients are often PMs and analysts who could become seat buyers, but many are only readers. Test how many shared-brief viewers start a trial, and whether a PM then invites their team.
- First customers
- Independent investment shops and small hedge fund research teams of 2 to 10 analysts, who already pay for data and use general AI tools by hand.
- Where to reach them
- Hacker News and finance dev forums: Share a worked example showing token use and context carry-over on an earnings update, and invite readers to a waitlist.
- Direct outreach on LinkedIn: Message analysts and PMs at small shops, offering a free pilot on one real coverage name.
- Investing communities: Post public sample workspaces on well-known tickers in analyst-focused groups and newsletters.
- First milestone
- Five small research teams pay for a month of use after a pilot, and at least three reopen the same workspace after a new earnings report.
Why people already pay for this
- VerifiedHacker News post received 148 points and 53 comments
- Public claimMCP tools waste tokens on financial data: one tool call for five years of daily prices dumps tens of thousands of tokens into the context window.
- Public claimFinancial tool schemas consume massive token overhead: data vendors pack dozens of tools into a single MCP server with schemas that can eat 50k+ tokens before the agent does anything useful.
- Public claimExisting agents lose research context across sessions: files don't carry over and users must re-paste context every time when updating analysis.
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Persistent sandboxIsolated code sandbox with per-workspace file storageAgents need to run code and keep files and memory between sessions without putting data in the context.
- This app · LLM providerHosted LLM API with tool useRuns the research agent and the generation of typed tool modules.
RisksAI inference
- DataData vendor licences may restrict storing, caching or sharing data in workspaces and exports. Many analysts may need to bring their own keys.
- AIAgents may misread figures or invent sources in financial work. Every claim must link to a source, and users must be able to check it.
- RegulatoryOutput used for investment decisions or shared with clients may raise compliance, record-keeping and advice concerns for regulated firms.
- CommercialLarger firms may require security reviews and may build this in-house. General agent tools may add persistent memory.
- TechnicalGenerating reliable typed modules from varied vendor APIs, and keeping a persistent sandbox cheap per user, is hard work.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.