Opportunities › AI-powered office document generation and automation › Proposed app
Deckhand
Turn a plain request into a slide deck, report or sheet, on models you choose
- Score
- 54
- confidence 45
- Evidence
- 4
- items on the page
- Rivals
- 2
- named in sources
- First version
- 35 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: Knowledge workers create documents at desks using rich text, templates, and brand kits in longer editing sessions. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
New document
Ready to generate- FormatPPTX
About this appAI proposal
Deckhand turns a written request into a formatted PowerPoint, Word, Excel or HTML file. Type what you need, attach notes or data, and get a file that opens in the tools you already use. It installs as a PWA.
You choose the model. Connect your own key for Claude, DeepSeek or Qwen, or point it at a model running on your own machine through Ollama. Prompts and files go only where you send them, so teams can keep sensitive material away from third-party document services.
Deckhand is for people who make the same kinds of documents again and again: weekly reports, client decks, price sheets. Save a request as a recipe, run it on new inputs, and pay your model provider's rate directly.
- Tagline
- Say what the document is. Get the file. Keep your data.
- Voice
- PlainPracticalCalm
- Why this name
- A deckhand does the routine work on a ship so the crew can steer. It also hints at decks of slides.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Plain request to real files
Describe the document. Get PPTX, DOCX, XLSX or HTML with headings, tables and layouts, not a text dump to paste.
No member yet
Bring your own model
Use your own API key for Claude, DeepSeek or Qwen, or a local Ollama model. Switch per document.
No member yet
Local-first option
Run with a local model so prompts and source files stay on your machine or network.
No member yet
Reusable recipes
Save a request as a recipe, then rerun it on new notes or data for the next weekly report or deck.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#0F4C4ADeep Teal · primary#F6F2EAPaper · surface#1D2625Ink · text#E2692BSignal Orange · accent#8DB5A5Sage · data
Typography
Aa
Deckhand
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
Say what the document is. Get the file. Keep your data.
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, based on the idea of free self-hosting with paid team features and support.
Self-host
Free
- Bring your own key or local model
- All four file formats
- Recipes and brand kit
- Single user
Team
$12 a month
- Per seat, shared recipes
- Shared brand kits
- Usage and cost dashboard
- Email support
Business
Contact us
- Single sign-on
- Audit log
- Install help and training
- Priority support
Would you pay Team · $12 a month?
Your answer tells PWAS which prices people accept. It changes no score.
No answers yet.
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
$194,000
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. Users send generated decks and reports to colleagues and clients. Files can carry an optional small credit, and shared HTML pages can link back to Deckhand.
- Do they become customers?
- Recipients see the output but may not be buyers. Test whether recipients who make similar documents click through and sign up, with the credit on and off.
- First customers
- Small consulting and operations teams of 5 to 30 people who produce weekly client reports and decks and have rules about sending data to cloud AI.
- Where to reach them
- Open-source communities: Post the self-host release in GitHub, Ollama and local-LLM forums, and answer document-automation questions there.
- Direct outreach: Message ops leads at small consultancies with a sample report built from their own public template.
- First milestone
- Five teams in one vertical running weekly recipes for a month, and two of them paying for Team seats.
Why people already pay for this
- VerifiedGitHub repository gained 321 stars in 67 days
- Public claimCloud-based AI office tools like Claude Cowork and Codex exist and are used for document generation, creating a market for document automation.
- Public claimThere is demand for alternatives to proprietary cloud solutions, evidenced by the project positioning itself as an open-source alternative and supporting multiple model backends (BYOK - Bring Your Own Key).
- Public claimPrivacy and cost control are concerns in the market, as the project emphasizes being 'local-first' and self-hosted.
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 · Document renderingOpen-source PPTX, DOCX and XLSX librariesBuilds real Office files from structured model output, on the server or in the browser.
- This app · Model connectorsProvider-neutral LLM client with Ollama supportLets users bring their own key or a local model without code changes.
RisksAI inference
- TechnicalSmaller local models may produce weak layouts or broken tables. Output quality needs a structured intermediate format and many test documents.
- CommercialOpen-source alternatives are free, so paying for a team tier needs a clear gain in setup, support or shared features.
- PlatformA PWA in the browser cannot always reach a local model server. Some users may need a local helper or companion install.
- Intellectual propertyOffice file fidelity depends on third-party libraries, and their licences must allow commercial use.
- AIModels can invent figures in generated reports, so users need source checks and review before sending.
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.