PanelPress
PanelPress turns raw CT slices into journal-ready, identifier-scrubbed figure panels in about a minute, for medical writers and radiology researchers racing a manuscript or poster deadline.
One product, one founder, thirty days. Everything below is built to be executed inside the contest window — the official rules still decide what counts.
Decide in 20 seconds
The facts that decide whether this is worth thirty days of yours.
Freelance or in-house medical writer, plus the radiology researcher who supplies the images, at a medical communications agency, CRO or academic imaging lab of 1-50 people in the US or EU, triggered by a manuscript revision or conference poster deadline.
Operational simplicity 3.3/10
Remove regulated data, human review steps and custom infrastructure; keep the product self-serve and stateless on managed services.
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 12; 53 sales at $19 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.0/10.
Score and projections come from the generated report and are not guarantees. The $100K Hackathon is run by Ship or Die — read the official rules before you build.
Your first 24 hours
12 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Freelance or in-house medical writer, plus the radiology researcher who supplies the images, at a medical communications agency, CRO or academic imaging lab of 1-50 people in the US or EU, triggered by a manuscript revision or conference poster deadline.
Why now
Deadline-driven publishing work has a discretionary card and a one-file deliverable, so value is provable in a 40-second demo with no integration, no procurement and no clinical or diagnostic claim.
Problem
Each CT figure costs one to three hours of screenshotting, overlay-border cropping, identifier blackout and PowerPoint panel layout, and one missed burned-in patient name is a privacy incident plus a submission rejection.
Solution
Upload a DICOM series or JPEG/PNG, pick the slice from a thumbnail strip, confirm OCR-flagged identifier blackout boxes, get an auto-cropped image with a scale bar from PixelSpacing, compose a 2-6 image labelled panel, and export a 300 DPI TIFF/PNG/PDF plus a redaction log PDF.
1. Product opportunity and positioning
Every oncology or radiology manuscript and poster carries CT figures, and nearly all of them are assembled by hand: screenshot the viewer, crop the scanner overlay border, black out burned-in patient names, rebuild a labelled panel in PowerPoint, export at 300 DPI, then write a sentence for the ethics statement about how identifiers were removed. PanelPress compresses that one-to-three-hour loop into roughly a minute of clicking.
The buyer is a named person, not a department: a medical writer at a medical communications agency or CRO, or the radiology researcher who hands them the images. The trigger is a dated deliverable, a manuscript revision or a poster deadline. Nothing in that purchase needs procurement, integration or a clinical claim. PanelPress measures nothing, segments nothing and diagnoses nothing. It makes a picture publishable and keeps a receipt of what was redacted.
JEV scored this concept 56/100 (band C). Its strongest signals were speed to first revenue (8.0) and reachable customers (6.8); the weakest were operational simplicity (3.3) and market demand (4.1). The plan is shaped by exactly that reading: sell fast to a buyer you can name by hand, keep the product stateless and self-serve, and never let it grow a human review step, a PACS integration or a regulated-data promise. Where market demand is uncertain, the answer is not a broader platform but a tighter deadline trigger and a demo that lands in forty seconds.
Positioning line: the fastest way to get a defensible CT figure out of a viewer and into a submission. The alternative is not a competitor product, it is a screenshot tool, PowerPoint, and a colleague squinting at a burned-in name at 11pm.
2. Smallest sellable MVP
One page, no account, no onboarding. Drag in a file, see a thumbnail strip, pick the slice, see flagged identifier boxes you must confirm, get an auto-cropped image with a scale bar, drop two to six images into a labelled panel, export a 300 DPI TIFF and a redaction log PDF, pay by credit pack. Anything that is not on that path is out of scope for the first six days: no PACS, no DICOMweb, no teams, no batch mode, no compliance certification, no diagnostic output.
3. Technical implementation plan
Next.js on Vercel, with DICOM parsing and window/level rendering in the browser so pixels never need a long-lived server process. Tesseract.js runs the overlay OCR pass client-side and dcmjs checks PatientName, PatientID, BirthDate and AccessionNumber. Composition and 300 DPI export use sharp and pdf-lib. Uploads land in object storage with a 24-hour lifecycle deletion rule. Stripe Checkout sells credits, a webhook writes the balance to Postgres, and every export decrements one credit. The hard part is DICOM transfer syntax variety, so the plan builds a plain JPEG/PNG fallback on day two and never relies on one parser path.
4. Pricing and monetization
Three one-time credit packs: $19 Single Panel, $99 Figure Pack (8 exports), $249 Poster Sprint (25 exports, poster presets, up to 3 seats). Blended target order value is $99. Credit packs beat subscriptions here because the work is intermittent and deadline-shaped; a subscription invites cancellation between manuscripts. The upsell is a second pack within the same poster season and a Poster Sprint when an agency ships a set.
5. First-customer acquisition
Sixty hand-built names is the whole channel. LinkedIn search for medical writer or scientific writer titles combined with oncology, radiology or imaging, plus public freelance directories and corresponding-author listings on recent imaging papers. Each message carries a forty-second screen recording and an offer to process one of their own files free within twelve hours, deleted automatically after twenty-four. When they ask for a second panel, the reply is a checkout link, not a call. Follow up on day three and day seven with a deadline-framed line.
6. Detailed 30-day execution roadmap
Days 1-5 are pure build: upload and slice picker, identifier scanner, crop and scale bar, panel composer, export and redaction log. Day 6 wires Stripe credits and deploys the MVP, which is also the build deadline. Day 7 records the demo and publishes the pricing page. Days 8-10 build and send the sixty-name list in three batches of twenty. Days 11-14 are free-panel delivery, follow-ups and removal of the top conversion blocker. Days 15-21 widen to research labs, pitch the Poster Sprint to agencies, and ship the three most-requested fixes. Days 22-30 are a second outreach round, reliability work on DICOM edge cases, refund hygiene, a final countdown push and a reconciled profit tally.
7. Revenue and profit scenarios (projections only)
Three projections, all modelled, none verified. Conservative: 3 customers at a $99 blended order value, $297 gross, no refunds, $12 processing fees, no ad spend, $25 operating cost. Base: 10 customers, $990 gross, $40 fees, $60 operating cost. Upside: 22 customers at a $110 blended value including Poster Sprints, $2,420 gross, $99 refunded, $109 fees, $120 operating cost. Paid marketing is zero in all three because the channel is organic outreach. All figures are scenario models from stated assumptions, not verified revenue, customers or traction.
8. Main risks and validation experiments
First, institutional security review stalls the impulse buy; test by counting how many of the first twenty ask for a DPA before uploading. Second, identifier OCR errors erode trust; test by running the scanner over ten synthetic slices with known burned-in text and measuring flag recall. Third, PowerPoint may be good enough; test by asking each free-panel recipient what they would have done instead. Fourth, DICOM variety breaks uploads; test with compressed and multi-frame samples before day six. Fifth, deadline-day responsiveness is one person's limit; mitigate with an FAQ, an email-only promise and a stated twelve-hour turnaround.
9. First 24-hour action plan
Hours 1-2: register the domain, create the repo, deploy a skeleton to Vercel, open a Stripe account in test mode. Hours 3-5: get DICOM parsing and window/level rendering working on one synthetic file. Hours 6-8: build the thumbnail strip and slice selection. Hours 9-12: start the identifier scanner, write the outreach message, and begin the first twenty rows of the sixty-name list so day nine is not blocked on research.
Smallest sellable MVP
A single-page, no-account tool that takes one CT file, lets the buyer confirm flagged identifier boxes, and returns a 300 DPI labelled figure panel plus a redaction log PDF after a Stripe credit purchase.
Must-have features
- Drag-and-drop upload for .dcm, zipped series and flat JPEG/PNG with soft-tissue, bone and lung window presets
- Thumbnail strip with arrow-key slice stepping
- OCR plus DICOM tag identifier scanner listing probable identifiers as one-click confirmable blackout boxes
- Auto crop-to-image-bounds removing the scanner overlay border
- Scale bar computed from PixelSpacing with editable unit label
- Panel composer for 2-6 images with A/B/C labels, arrows and text callouts
- 300 DPI TIFF/PNG/PDF export with optional caption block
- Redaction log PDF recording every flagged region and user action
- Stripe Checkout credit packs with automatic 24-hour file deletion
Explicitly out of scope
- No diagnostic measurement, lesion segmentation or reader functionality
- No PACS, DICOMweb or HL7 integration
- No team accounts, SSO, admin console or role permissions
- No compliance certification or guaranteed de-identification claim
- No batch processing of hundreds of series
- No support beyond email and an FAQ page
Demo flow (what the buyer sees)
- Founder drags a synthetic CT series into the browser
- Thumbnail strip appears; founder arrow-steps to the target axial slice
- Identifier scanner flags a burned-in fake name; founder clicks one blackout box to confirm
- Auto crop removes the overlay border and the scale bar renders from PixelSpacing
- Two slices are placed in a labelled A/B panel with an arrow callout
- Export produces a 300 DPI TIFF plus a redaction log PDF in about one minute
Technical implementation plan
Stack
Architecture
All pixel work happens in the browser tab: the file is parsed locally, rendered to canvas with a window/level preset, OCR'd locally, and composed locally into an export blob. The server only mints Stripe sessions, receives the webhook, writes a credit balance and stores the redaction log record; uploaded originals go to object storage with a lifecycle rule that deletes them after 24 hours so no long-lived patient pixel store exists.
Build steps
- Scaffold Next.js on Vercel with a deploy preview and a domain
- Implement DICOM parse, PixelSpacing extraction and canvas window/level rendering
- Add the thumbnail strip and exact slice selection
- Run Tesseract OCR over the rendered pixels and cross-check PatientName, PatientID, BirthDate and AccessionNumber tags
- Draw confirmable blackout boxes and write each decision to a redaction record
- Add auto crop-to-image-bounds and the PixelSpacing-derived scale bar
- Build the 2-6 image panel composer with labels, arrows and callouts
- Implement 300 DPI TIFF/PNG/PDF export and the redaction log PDF
- Wire object storage with a 24-hour deletion lifecycle
- Integrate Stripe Checkout credit packs and the webhook that credits the balance
- Ship a JPEG/PNG fallback path and a clear error state for unsupported transfer syntaxes
- Deploy the public MVP and record the forty-second demo
AI components
- Tesseract.js OCR over rendered pixel data to flag burned-in overlay text for user-confirmed blackout
- Rule-based DICOM tag scoring across PatientName, PatientID, BirthDate and AccessionNumber; no LLM in the paid path
Pricing and monetization
Single Panel
Entry offerFor Researcher or student testing the tool on one figure
- 1 panel export at 300 DPI
- Redaction log PDF
Figure Pack
For Medical writer preparing one manuscript or poster
- 8 panel exports
- Redaction log PDFs
- Scale bar and callout library
- Priority email replies
Poster Sprint
For Comms agency or imaging lab shipping a conference poster set
- 25 panel exports
- Poster-size layout presets
- Shared credit balance across up to 3 seats
Model: One-time credit packs purchased through Stripe Checkout, with one export consuming one credit and no subscription.
First revenue offer: Send the demo link plus: I will process one of your CT files into a labelled 300 DPI panel free within 12 hours, your file auto-deletes after 24; if you want a second panel, the $99 Figure Pack link is right here.
Expected ACV: $99
Upsell path: A writer who uses the Figure Pack on one manuscript buys a second pack in the next deadline cycle, and an agency contact who ships poster sets moves to the $249 Poster Sprint with three seats on one credit balance.
First-customer acquisition
Cold LinkedIn and email outreach to 60 hand-built medical writer names, each message carrying a forty-second screen recording.
Where to find them
- LinkedIn search for medical writer / scientific writer titles combined with oncology, radiology or imaging
- Public freelance medical writing and scientific editing directories
- Corresponding-author and first-author listings on recent open-access imaging papers
- Medical communications agency team pages and CRO medical writing rosters
- Radiology and imaging research lab pages at academic centres
Exact outreach motions
- Build 60 rows with name, title, employer, profile URL and the last paper or poster they authored
- Send batch 1 of 20 with the demo link and the free-panel offer, three sentences maximum
- Send batch 2 of 20 the next day with the same message, no variation
- Send batch 3 of 20 and log every reply in a single sheet with stage and blocker
- Follow up on day 3 and day 7 with a deadline-framed line asking for their manuscript or poster date
- When a free panel lands, ask for the second file and send the $99 checkout link
Message template
Hi {first_name} - I saw your {month} imaging manuscript. I built PanelPress: drop in a CT slice, confirm the flagged patient-name boxes, get a labelled 300 DPI panel plus a redaction log for the ethics statement, about a minute. 40-second demo: {link}. Send me one of your files and I will do the first panel free within 12 hours; it auto-deletes after 24. Second panel is a $99 pack if it is useful.30-day execution roadmap
Week 1 · Build & validate · 7 days
Ship a deployed shell
- Register the domain and create the Next.js repo on Vercel.
- Open Stripe in test mode and set the product env vars.
- Deploy a skeleton page with a working upload dropzone that stores nothing.
Deliverable: Live skeleton URL with a dropzone and test-mode Stripe keys.
Get pixels on screen
- Parse .dcm and zipped series with dcmjs and read PixelSpacing.
- Render to canvas with soft-tissue, bone and lung window presets.
- Add a JPEG/PNG fallback path for non-DICOM files.
Deliverable: Browser renders a synthetic CT slice with three presets.
Exact slice selection
- Build the thumbnail strip from the series.
- Add arrow-key stepping and a selected-slice indicator.
- Handle large series without freezing the tab.
Deliverable: Thumbnail strip with keyboard slice stepping.
The trust feature
- Run Tesseract OCR over the rendered pixels.
- Check PatientName, PatientID, BirthDate and AccessionNumber tags.
- Draw confirmable blackout boxes and record each user decision.
Deliverable: Flag list plus one-click confirmable blackout boxes.
The paid artifact
- Add auto crop-to-image-bounds and the PixelSpacing scale bar.
- Build the 2-6 image panel composer with A/B/C labels and arrows.
- Export 300 DPI TIFF/PNG/PDF plus the redaction log PDF.
Deliverable: Full panel export and redaction log PDF from one file.
MVP deadline
- Wire Stripe Checkout credit packs and the balance webhook.
- Enforce one credit per export and 24-hour object deletion.
- Deploy the public MVP with pricing on the page.
Deliverable: Public MVP accepting live payments at panelpress pricing.
Make outreach possible
- Record the forty-second slice-to-panel screen capture.
- Publish the landing page with three tiers and an FAQ.
- Write the removal-and-redaction explainer so no compliance claim is made.
Deliverable: Demo video, live pricing page and FAQ.
Week 2 · Launch & first sale · 7 days
Sixty named buyers
- Build the first 30 rows with name, title, employer and last paper.
- Add 30 more from freelance directories and author listings.
- Verify each title is writer, editor or imaging researcher.
Deliverable: 60-row outreach sheet with profile URLs.
First 20 sends
- Send 20 messages with the demo link and free-panel offer.
- Set a 12-hour free-panel turnaround rule for yourself.
- Log every send and reply in the sheet.
Deliverable: 20 sends logged, replies tracked.
Second 20 sends
- Send the next 20 with identical copy.
- Add reply-triage statuses to the sheet.
- Fix any upload error reported by the first replies.
Deliverable: 40 total sends and a working triage column.
Earn the second file
- Deliver every received file as a panel within 12 hours.
- Ask each recipient what they would have done without the tool.
- Record the top objection verbatim.
Deliverable: Delivered panels plus a verbatim objection list.
First paying customer
- Send day-3 and day-7 follow-ups with deadline framing.
- Send the $99 Figure Pack link to anyone asking for a second panel.
- Watch Stripe for the first live payment.
Deliverable: First paid credit pack, or a logged reason it did not land.
Complete the 60
- Send the final 20 messages.
- Reply to every open thread within the same day.
- Offer the $19 Single Panel to anyone hesitating on price.
Deliverable: 60 sends complete with reply-rate number.
Kill the top objection
- Publish a one-page data-handling note covering upload, deletion and redaction sign-off.
- Add an inline note that the user confirms every redaction.
- Re-test the checkout on mobile.
Deliverable: Data-handling page linked from the pricing block.
Week 3 · Repeat & systematize · 7 days
Poster buyers
- Record a poster-set demo showing the layout preset.
- Add poster-size preset options to the composer.
- Email the demo to the three most engaged replies.
Deliverable: Poster demo video and preset options shipped.
$249 tier
- Identify 15 medical communications agencies and CRO writing teams.
- Send a Poster Sprint pitch to each with the poster demo.
- Offer a shared 3-seat credit balance explicitly.
Deliverable: 15 agency pitches sent, Poster Sprint page live.
Researcher side of the pair
- List 20 imaging lab PIs and research fellows from recent papers.
- Message with the figure-prep angle, not a clinical angle.
- Ask each whether a writer or they assemble the figures.
Deliverable: 20 lab contacts and a map of who owns the figure.
Convert the warm list
- Follow up with everyone who received a free panel.
- Log the top three reasons for not buying.
- Refund anyone who asks, without friction.
Deliverable: Objection summary and any conversion from the warm list.
Top three asks
- Implement the most requested export or crop change.
- Improve the error message for unsupported DICOM syntax.
- Re-run the full demo path end to end.
Deliverable: Top three requested fixes deployed.
Publish the method
- Publish a short page on exactly what the scanner checks and what it cannot catch.
- State plainly that the user signs off on every redaction.
- Link the redaction log sample PDF.
Deliverable: Public identifier-handling page with a sample log.
Referral ask
- Ask every buyer for one colleague name.
- Compute the reply rate, free-panel rate and paid rate.
- Decide which single segment gets week four.
Deliverable: Funnel numbers and a chosen segment for week 4.
Week 4 · Scale what works · 7 days
$19 tier
- Pitch the Single Panel to fellows and students in the list.
- Measure whether the low tier pulls in first-time buyers.
- Keep the Figure Pack as the default recommendation.
Deliverable: Entry-tier pitch sent to 15 low-budget contacts.
Non-responders
- Re-message 20 non-responders with a new subject line.
- Reference a real deadline month, not a generic urgency claim.
- Log any reply within the hour.
Deliverable: 20 second-touch messages sent.
Make the $249 tier real
- Ship shared credit balance across up to three seats.
- Add poster-size canvas presets to the composer.
- Write the Poster Sprint checkout description.
Deliverable: Working 3-seat Poster Sprint purchase.
Permission first
- Ask to post a build note in two research communities you already belong to.
- Answer questions in-thread rather than pitching.
- Track referral clicks from those posts.
Deliverable: Two community posts with tracked referral traffic.
Highest-value deals
- Follow up on every agency pitch.
- Offer to process their next poster set as a paid Poster Sprint.
- Escalate any that stall to a same-day email reply.
Deliverable: Agency pipeline status for every pitch.
Stop the leaks
- Add a clearer unsupported-file error with a JPEG workaround.
- Test five DICOM export variants from different scanners.
- Fix the highest-frequency upload failure.
Deliverable: DICOM failure rate reduced on a five-file test set.
Clean books
- Reconcile Stripe charges against credited exports.
- Check every refund and confirm no chargeback is pending.
- Verify object storage deletion is actually running.
Deliverable: Reconciled credit and payment ledger.
Days 29-30 · Consolidate · 2 days
Last deadline framing
- Send a final outreach round to the entire 60-name list.
- Push the Poster Sprint to any unconverted agency thread.
- Watch for and handle refunds same-day.
Deliverable: Final outreach round sent and logged.
Verified numbers only
- Compute collected revenue minus refunds, chargebacks and fees.
- Compare actuals to the conservative and base projections.
- Write the post-mortem and the next-30-day decision.
Deliverable: Verified profit tally and a written next-step decision.
Revenue and profit scenarios
How the numbers are defined
Hackathon profit — Collected revenue − refunds − chargebacks − payment processing fees − paid marketing spend.
Business profit — Hackathon profit − operating costs (hosting, domains, tooling, AI/API delivery costs).
The contest ranks entrants on hackathon profit only. Hosting, tooling and AI/API delivery costs are not deducted by the organisers, but each sale still has to cover them — so we show both. Definitions verified on 2026-11-15 against the official rules.
Stated assumptions
- All three scenarios are projections from stated assumptions, not verified revenue or existing customers.
- Blended average order value is $99 (Conservative, Base) and $110 (Upside) reflecting a mix of $19, $99 and $249 packs.
- Payment fees assume Stripe Checkout at 2.9% + $0.30 per transaction, modelled at roughly 4-4.5% blended to cover international cards.
- Paid marketing spend is $0 in every scenario because the primary channel is organic cold outreach and free-panel delivery.
- Operating cost covers the domain, Vercel, object storage with 24-hour deletion, and OCR/export compute; OCR runs client-side so inference cost stays near zero.
- Conservative assumes 3 buyers from the first 60 conversations.
- Base assumes 10 buyers, roughly a 17% close rate on 60 targeted conversations.
- Upside assumes 22 buyers including two Poster Sprints and one $99 refund.
- 40 paying customers in 30 days from the concept math is treated as unreached by these scenarios and is not claimed.
Conservative
3 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $12
- Paid marketing
- $0
- Hackathon profit
- $285
- Operating costs
- $25
- Business profit
- $260
Base case
10 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $40
- Paid marketing
- $0
- Hackathon profit
- $950
- Operating costs
- $60
- Business profit
- $890
Optimistic
22 customers- Refunds
- $99
- Chargebacks
- $0
- Payment fees
- $109
- Paid marketing
- $0
- Hackathon profit
- $2,212
- Operating costs
- $120
- Business profit
- $2,092
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $190
- Payment fees (3.9%)
- $7
- Paid marketing
- $0
- Hackathon profit
- $183
- Business profit
- $123
Hypothetical arithmetic, not a forecast. The fee rate is a flat 3.9% assumption, refunds and chargebacks are left at zero here, and none of this is tracked data.
Scenario models from stated assumptions, not verified revenue. No customers, partnerships or traction exist at the start date.
Main risks and validation experiments
Risk 1Buyers demand an institutional security review or a signed data processing agreement before uploading any file, which kills the impulse purchase.
State plainly that files auto-delete in 24 hours, publish the data-handling page, and keep the offer to one file so no institutional process is required.
Count how many of the first 20 replies ask for a DPA or security questionnaire before sending a file.
If more than half of the first 20 engaged prospects require institutional approval before uploading, reposition toward a no-upload local-only tool or switch the buyer.
Risk 2OCR false positives and false negatives erode trust in the identifier flag and create manual re-checking that costs more time than PowerPoint.
Frame the scanner as an assistant the user signs off on, never as de-identification, and show every flag with its source (pixel OCR or DICOM tag).
Run the scanner over ten synthetic slices with known burned-in text and known clean overlays, and record flag recall and false-flag count.
If recall on the ten-slice test set is below roughly 70%, stop selling the identifier feature as a benefit and sell the panel assembly alone.
Risk 3PowerPoint templates and viewer export buttons are considered good enough, so the three-hour pain is tolerated rather than solved.
Lead the demo with the redaction log and the scale bar, the two things PowerPoint does not produce, and price entry at $19.
Ask every free-panel recipient what tool they used on their last figure and how long it took, and log the verbatim answer.
If ten free-panel recipients report under 30 minutes per figure and none asks for a second panel, the pain is too small and the wedge needs replacing.
Risk 4DICOM transfer syntax variety breaks uploads on real buyer files during a live demo or deadline.
Ship a JPEG/PNG fallback on day two, add a clear unsupported-file error with a manual workaround, and test five scanner variants before day 27.
Collect five real DICOM exports from different sources and run them through the full path end to end.
If fewer than three of five real files complete the full export path by day 27, stop outreach and fix ingestion before the final week.
Risk 5A single founder cannot match deadline-day responsiveness, so refunds land at the worst moment and the redaction log is never used.
Promise twelve-hour turnaround, not instant support, publish an FAQ, and handle every refund same-day without argument.
Time every free-panel delivery in week two and record the worst case.
If worst-case delivery exceeds 24 hours twice in week two, cut the free-panel offer and switch to demo-only outreach.
Hour-by-hour: the first 24 hours
Hours 1-2
- Register the panelpress domain and create the Next.js repo.
- Connect the repo to Vercel and deploy the default page.
- Open a Stripe account in test mode.
Output: Live skeleton URL and test-mode Stripe keys.
Hours 3-5
- Install dcmjs and wire a file input that parses a .dcm file.
- Extract PixelSpacing and render the pixel array to canvas.
- Download one public synthetic CT sample for testing.
Output: A canvas rendering of a parsed CT slice in the browser.
Hours 6-8
- Add soft-tissue, bone and lung window/level presets.
- Build the thumbnail strip from series frames.
- Add arrow-key slice stepping.
Output: Working slice picker with three presets.
Hours 9-12
- Start the Tesseract.js overlay OCR pass.
- Write the three-sentence outreach message and the free-panel offer.
- Fill the first 20 rows of the 60-name sheet.
Output: OCR prototype, final message copy and 20 verified outreach rows.
Full 30-day launch report
Read the full generated report
PanelPress — 30-Day Launch Report
1. Product opportunity and positioning
Every oncology or radiology manuscript and poster carries CT figures, and nearly all of them are assembled by hand: screenshot the viewer, crop the scanner overlay border, black out burned-in patient names, rebuild a labelled panel in PowerPoint, export at 300 DPI, then write a sentence for the ethics statement about how identifiers were removed. PanelPress compresses that one-to-three-hour loop into roughly a minute of clicking.
The buyer is a named person, not a department: a medical writer at a medical communications agency or CRO, or the radiology researcher who hands them the images. The trigger is a dated deliverable, a manuscript revision or a poster deadline. Nothing in that purchase needs procurement, integration or a clinical claim. PanelPress measures nothing, segments nothing and diagnoses nothing. It makes a picture publishable and keeps a receipt of what was redacted.
JEV scored this concept 56/100 (band C). Its strongest signals were speed to first revenue (8.0) and reachable customers (6.8); its weakest were operational simplicity (3.3) and market demand (4.1). The plan is shaped by that reading: sell fast to a buyer you can name by hand, keep the product stateless and self-serve, and never let it grow a human review step, a PACS integration or a regulated-data promise. Where demand is uncertain, the answer is a tighter deadline trigger and a demo that lands in forty seconds, not a broader platform.
The alternative is not a competitor product. It is a screenshot tool, PowerPoint, and a colleague squinting at a burned-in name at 11pm.
2. Smallest sellable MVP
One page, no account, no onboarding. Drag in a file, see a thumbnail strip, pick the slice, confirm flagged identifier boxes, get an auto-cropped image with a scale bar, drop two to six images into a labelled panel, export a 300 DPI TIFF and a redaction log PDF, pay by credit pack. Everything else is deferred: no PACS, no DICOMweb, no teams, no batch mode, no certification, no diagnostic output.
3. Technical implementation plan
Next.js on Vercel, with DICOM parsing and window/level rendering in the browser so pixels never need a long-lived server process. Tesseract.js runs the overlay OCR pass client-side and dcmjs checks PatientName, PatientID, BirthDate and AccessionNumber. Composition and 300 DPI export use sharp and pdf-lib. Uploads land in object storage with a 24-hour deletion lifecycle. Stripe Checkout sells credits, a webhook writes the balance to Postgres, and every export decrements one credit. The hard part is DICOM transfer syntax variety, so a plain JPEG/PNG fallback ships on day two and the product never depends on one parser path.
4. Pricing and monetization
Three one-time credit packs: $19 Single Panel, $99 Figure Pack with eight exports, $249 Poster Sprint with twenty-five exports, poster presets and up to three seats. Blended target order value is $99. Credit packs beat subscriptions because the work is intermittent and deadline-shaped; a subscription invites cancellation between manuscripts. The upsell is a second pack in the next deadline cycle, and a Poster Sprint when an agency ships a poster set.
5. First-customer acquisition
Sixty hand-built names is the whole channel. LinkedIn search for medical writer or scientific writer titles combined with oncology, radiology or imaging, plus public freelance directories, agency team pages and corresponding-author listings on recent imaging papers. Each message carries a forty-second screen recording and an offer to process one of their own files free within twelve hours, deleted automatically after twenty-four. When they ask for a second panel, the reply is a checkout link, not a call. Follow up on day three and day seven with a deadline-framed line.
6. Detailed 30-day execution roadmap
Days 1-5 are pure build: upload and slice picker, identifier scanner, crop and scale bar, panel composer, export and redaction log. Day 6 wires Stripe credits and deploys the MVP, which is also the build deadline. Day 7 records the demo and publishes pricing. Days 8-10 build and send the sixty-name list in three batches of twenty. Days 11-14 are free-panel delivery, follow-ups and removal of the top conversion blocker. Days 15-21 widen to research labs, pitch the Poster Sprint to agencies and ship the three most-requested fixes. Days 22-30 are a second outreach round, reliability work on real DICOM files, refund hygiene, a final countdown push and a reconciled profit tally.
7. Revenue and profit scenarios (projections only)
Three projections, all modelled, none verified. Conservative: 3 customers at a $99 blended order value, $297 gross, no refunds, $12 in processing fees, no ad spend, $25 operating cost. Base: 10 customers, $990 gross, $40 fees, $60 operating cost. Upside: 22 customers at a $110 blended value including Poster Sprints, $2,420 gross, $99 refunded, $109 fees, $120 operating cost. Paid marketing is zero in all three because the channel is organic outreach. All figures are scenario models from stated assumptions, not verified revenue, customers or traction.
8. Main risks and validation experiments
First, institutional security review stalls the impulse buy; test by counting how many of the first twenty ask for a DPA before uploading. Second, OCR errors erode trust; test on ten synthetic slices with known burned-in text and measure flag recall. Third, PowerPoint may be good enough; ask each free-panel recipient what they used last time and how long it took. Fourth, DICOM variety breaks uploads; test five real scanner exports before day 27. Fifth, deadline-day responsiveness is one person's limit; mitigate with twelve-hour turnaround, an FAQ and same-day refunds.
9. First 24-hour action plan
Hours 1-2: domain, repo, Vercel deploy, Stripe test account. Hours 3-5: parse a .dcm file, read PixelSpacing and render to canvas. Hours 6-8: window presets, thumbnail strip, arrow-key stepping. Hours 9-12: start the OCR pass, finalise the outreach message and fill the first twenty rows of the sixty-name sheet so day nine is not blocked on research.
This brief was generated from public data about established startups, then narrowed and scored for a solo 30-day launch. Revenue and profit figures are scenario models, not verified results, guarantees or evidence of existing customers. Validate demand yourself before you spend a month on it — and check the contest rules for what counts toward your score.
Related ideas
All ideas →ReviewReady Packet
Generates a reviewer-ready Meta App Review submission bundle — per-permission justifications, screencast script, test-account steps and deauth code — for solo devs and small agencies stuck at 'rejected'.
Chargeback Brief
Chargeback Brief turns a Stripe or Shopify Payments dispute into a bank-ready evidence packet — cover letter, delivery timeline and exhibit index — in under ten minutes for Shopify stores that currently lose disputes by missing the response deadline.
Flowproof Audit
Paste in an n8n or Make workflow export and get a scored reliability report plus drop-in patch nodes that stop silent failures, duplicate runs and untyped AI outputs before a client notices.