DocketDraft
DocketDraft turns one proposed federal rule into a submission-ready public comment letter plus a page-anchored citation ledger, built for solo public-affairs consultants and small advocacy nonprofits working against a filing 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.
Solo public-affairs and policy consultants, and the single policy person at advocacy nonprofits, coalitions and state-level associations with 1-10 staff, who already file comments on US federal rulemakings.
Monetization potential 4.9/10
Tie the price to a measurable money or time outcome and add a recurring or repeat-purchase path instead of a single low-value charge.
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 11; 5 sales at $249 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.2/10.
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
11 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Solo public-affairs and policy consultants, and the single policy person at advocacy nonprofits, coalitions and state-level associations with 1-10 staff, who already file comments on US federal rulemakings.
Why now
Federal comment windows are short, public and deadline-driven, and the commenters are named in public docket threads, so a first paying customer is reachable by direct message within days without ads, procurement or integration.
Problem
A relevant proposed rule opens with 10-35 days left, and one person must read 80-300 pages, pick the provisions that hit their members, and produce a cited letter before the deadline - so they miss the window or file two generic paragraphs.
Solution
Paste the docket URL or upload the proposed rule PDF, answer five questions about who you represent and which provisions matter, and receive a formatted DOCX/PDF comment letter plus a CSV citation ledger mapping every claim to a verbatim quote, page and section.
1. Product opportunity and positioning
DocketDraft sells one artifact: a submission-ready public comment letter plus a citation ledger that pins every argument to a quoted line, its page and its section in a proposed federal rule. The buyer is a solo public-affairs consultant, or the single policy person inside a 1-10 person advocacy nonprofit, coalition or state association. The trigger is not a budget cycle - it is a docket opening with roughly 10-35 days left on the clock.
That trigger is the whole commercial argument. The buyer is already committed to filing; the deadline creates urgency without any education budget; and the deliverable either exists or does not, so value is demonstrable on a live docket before a dollar changes hands. There is no procurement cycle, no integration, no data licensing and no regulatory approval in the path to the first payment.
The alternative set is narrow and bad. A general chat assistant produces fluent paragraphs and invented section numbers, which is a liability in a document an agency will read and opponents may quote back. Outside counsel produces a defensible letter for more than a small engagement earns. Doing it by hand is the status quo, and it costs the buyer a week they do not have.
Positioning is therefore concrete rather than aspirational: not a monitoring platform, not a legislative tracker, not a dashboard. DocketDraft is a single-artifact generator for the comment-letter deadline, sold one letter at a time. The narrowness is the point - the founder can hand-deliver a first issue section for a live docket in under 24 hours, which is impossible for a platform sale.
The concept was scored by JEV at 64.3/100 (band C), strongest on speed to first revenue (8.19) and reachable customers (7.8), and weakest on monetization potential (4.87) and operational simplicity (4.9). This 30-day plan attacks those two weak criteria directly: a repeat-purchase season pass is built into the pricing from day one rather than bolted on later, and the pipeline stays stateless on managed services with no human review step, no stored client data beyond the uploaded rule and no custom infrastructure.
Scope discipline is deliberate. Federal agencies only, proposed rules only, English only, export-only. Every excluded feature - monitoring, alerts, state legislation, hearings, filing into regulations.gov, collaboration - is excluded because it adds an operational surface without adding a single dollar to the first 30 days of collected revenue.
Smallest sellable MVP
A paid, single-flow web app: paste a docket URL or upload a proposed-rule PDF, answer five stance questions, pay $249 by Stripe Checkout, and receive a DOCX letter plus a CSV citation ledger by email - with a watermarked free preview of issue section 1 shown before payment.
Must-have features
- PDF or regulations.gov URL intake that preserves page markers in extracted text
- Section segmenter storing page and section offsets for every block
- Quote-lock extractor that builds a verbatim quote pool; drafting may cite only pool IDs
- Five-question stance intake (who you represent, provisions of concern, requested change, signatory, tone)
- Letter composer producing executive summary plus 3-5 numbered issue sections bound to quote IDs
- Citation ledger export as CSV plus an in-letter footnote table (quote, page, section, supported claim)
- DOCX and PDF export with the user's name, organization and signatory block
- Stripe Checkout one-time purchase gating the full letter; watermarked preview otherwise
- Resend delivery email with both files attached
Explicitly out of scope
- No crawling, watchlists, alerts or email digests of new rules
- No state legislative coverage, hearings, meeting transcripts or enforcement actions
- No direct filing into regulations.gov - the product ends at the exportable artifact
- No team seats, collaboration, comment threads or approval workflows
- No legal advice positioning or attorney review layer
- No stance recommendation - the user supplies the position
- No scanned-PDF OCR in v1 unless extraction failures block paying work
Demo flow (what the buyer sees)
- Founder opens the live app and pastes an open docket URL from regulations.gov
- App extracts text with page markers and shows segmented sections
- Founder answers the five stance questions for a sample association
- App builds the quote pool and displays selectable page-anchored quotes
- App drafts the letter and renders a watermarked preview of issue section 1
- Stripe Checkout completes the $249 purchase
- User receives DOCX letter and CSV citation ledger by email within two minutes
- User edits stance answers once and regenerates at no extra charge
Technical implementation plan
Stack
Architecture
A single Next.js app on Vercel holds every route. Intake resolves either a PDF upload or a regulations.gov docket number, downloads or receives the file, and runs pdfjs extraction that emits a text block array with page and character offsets. The segmenter sends that array to the long-context model and receives section boundaries as offsets, which are stored in Supabase. The quote-lock extractor runs per section and returns verbatim quotes with their offsets; the server re-validates each returned quote against the source text by string match and drops any quote that does not match exactly, so the pool contains only strings that exist in the rule. Drafting is a second model call that receives the stance answers plus the quote pool and may only emit quote IDs, which the server expands back into quotes, pages and sections when rendering the letter and the ledger. Stripe Checkout is created after the preview render and the webhook flips the order to paid, at which point DOCX/PDF generation and the Resend delivery fire. Nothing long-running, no queue, no human in the loop.
Build steps
- Day 1: repo, Vercel project, Supabase schema (dockets, segments, quotes, letters, orders), domain
- Day 2: PDF upload plus regulations.gov fetch, page-preserving extraction, segmenter endpoint
- Day 3: quote-lock extractor with server-side verbatim validation against source text
- Day 4: stance intake form, letter composer prompt, ledger generation, DOCX/PDF render
- Day 5: Stripe Checkout, webhook gating, watermarked preview, Resend delivery
- Day 6: production deploy, live keys, TrustMRR-connected payment account, landing page and one public demo letter
- Day 13: OCR fallback for scanned PDFs if extraction failures appear in real dockets
- Day 16: saved organization profile and season-pass entitlement (5 letters per month)
- Day 20: caching of extraction results to cut per-letter model cost
AI components
- Long-context segmentation of the full rule into numbered sections with offsets
- Per-section verbatim quote extraction into a locked pool
- Letter drafting constrained to quote-pool IDs, never free-text citations
- Optional single regeneration pass after the user edits stance answers
Pricing and monetization
Single Docket Letter
Entry offerFor Solo policy consultant or small nonprofit filing one comment before a deadline
- Full comment letter in DOCX and PDF
- Citation ledger CSV with page and section anchors
- One regeneration pass after the user edits the stance answers
Comment Season Pass
For Consultants or associations filing several comments in one quarter
- Up to 5 docket letters per month
- Citation ledger exports
- Saved organization profile and signatory block
Model: One-time per-letter purchase with an optional monthly season pass for repeat filers.
First revenue offer: For the live docket you name: I will draft issue section 1 with page-anchored quotes and send it as a PDF free within 24 hours, no signup. If it is usable, the full letter plus citation ledger is $249 before the comment window closes.
Expected ACV: $249
Upsell path: Single-letter buyers who file again within 60 days are offered the $399 season pass on their second docket; consultants who resell the artifact to multiple clients move to the pass as a standing tool rather than a per-filing cost.
First-customer acquisition
Direct one-to-one outreach to named commenters and solo policy consultants identified in public docket comment threads for currently open federal rulemakings.
Where to find them
- regulations.gov comment threads on recently closed related rules - the organizations and individuals who filed are named
- regulations.gov open dockets filtered to food and labeling, small-business labor, and energy permitting
- State restaurant, retail and manufacturing association staff directories (executive director and policy director pages)
- LinkedIn search for 'policy consultant' plus a state name, filtered to 1-10 person firms
- Trade-press comment-quote roundups that name the person who filed on a docket
- Government-affairs professional listservs and Slack communities where members post open dockets
Exact outreach motions
- Send a 90-word message naming the exact open docket number and its close date, offering one free drafted issue section within 24 hours
- Deliver the free section as a branded PDF with the citation footnote visible, and ask one question: which provisions did I miss
- Two days later, send the full letter skeleton status and the remaining hours on the clock
- Ask every buyer for one referral to another organization that commented on the same docket, offering the same free section
- Harvest newly opened dockets every Monday and Thursday and add five named prospects per docket before any other work
Message template
Hi [Name] - you filed on [prior rule] for [org]. [Agency] opened [docket no.] on [topic] and comments close [date]. I build comment letters that pin every claim to a quoted page and section of the proposed rule. Send me nothing: I will draft issue section 1 for [org] against that docket and email it as a PDF within 24 hours, free. If it is usable, the full letter plus the citation ledger is $249 before the deadline. Want it?
30-day execution roadmap
Week 1 · Build & validate · 7 days
Ship a deployed skeleton
- Create Next.js repo, Vercel project and docketdraft.com domain.
- Create Supabase tables for dockets, segments, quotes, letters and orders.
- Write the fixed letter structure spec (addressee, summary, 3-5 issues, requested changes, closing).
Deliverable: Deployed app with a health route and a committed letter spec.
Page-accurate text from real rules
- Build PDF upload and regulations.gov docket fetch.
- Extract text with pdfjs preserving page and character offsets.
- Test on three real proposed rules from different agencies.
Deliverable: Extraction endpoint returning page-marked blocks for three real dockets.
Only citable text enters the pool
- Prompt the long-context model for section boundaries as offsets.
- Extract per-section verbatim quotes and validate each by exact string match.
- Drop and log any quote that fails validation.
Deliverable: Quote pool JSON for three sections with zero unvalidated quotes.
Letter plus ledger from one run
- Build the five-question stance intake form.
- Draft the letter using quote IDs only and render the citation ledger.
- Produce DOCX and PDF with a signatory block.
Deliverable: Complete letter plus CSV ledger generated end to end from one docket.
Paid flow, watermarked preview
- Wire Stripe Checkout and the purchase webhook.
- Gate full output behind payment and render a watermarked issue-section-1 preview.
- Send DOCX and CSV by Resend after webhook confirmation.
Deliverable: End-to-end paid test purchase in Stripe test mode.
Live and chargeable
- Deploy with live Stripe keys and a payment account connected to TrustMRR.
- Publish the landing page and one public demo letter on a real docket.
- Post the demo in two government-affairs communities.
Deliverable: Production URL accepting real payments with a public demo letter.
60 named prospects
- Harvest named commenters from three live dockets.
- Add association staff and solo consultants from directories and LinkedIn.
- Send the first ten outreach messages.
Deliverable: Prospect sheet with 60 named contacts and 10 messages sent.
Week 2 · Launch & first sale · 7 days
Volume with specificity
- Send 15 messages, each naming a specific docket and close date.
- Deliver two free issue sections manually if the pipeline is slow.
- Log every reply and objection in the sheet.
Deliverable: 15 sends, 2 free sections delivered.
Second wave and follow-ups
- Send 20 more messages to newly listed prospects.
- Follow up on day-7 batch with the remaining-hours angle.
- Ask one warm replier which provisions matter most to their members.
Deliverable: 20 sends and 10 follow-ups logged.
Turn a preview into a payment
- Send the first Stripe checkout link to a replier with a live docket.
- Fix any draft defect the buyer flags the same day.
- Deliver the free section within 24 hours of every request.
Deliverable: First checkout link sent to a named buyer.
Close and deliver
- Close the first paid letter and deliver DOCX plus ledger.
- Confirm the payment lands inside the contest window.
- Record real model and hosting cost for that letter.
Deliverable: First collected $249 payment and delivered artifact.
Repeat the same motion
- Send 20 messages and follow up on every open thread.
- Deliver all pending free sections before any build work.
- Rewrite the drafting prompt with the failure notes from letters one and two.
Deliverable: Two more letters delivered or in draft.
Kill extraction failures
- Add OCR fallback for scanned PDFs.
- Add a clear failure message plus manual-upload path when extraction confidence is low.
- Log which agencies produce the worst PDFs.
Deliverable: OCR fallback deployed and failure logging live.
Reconcile and sweep
- Send 20 messages to dockets closing in the next 14 days.
- Send every 48-hour deadline reminder to open threads.
- Review paid, refunded and lost counts against the base scenario.
Deliverable: Midpoint numbers table and refreshed prospect list.
Week 3 · Repeat & systematize · 7 days
Second buyer per docket
- Ask every buyer for one referral to another commenter on the same docket.
- Offer the referred organization the same free issue section.
- Record which referral source produced a reply.
Deliverable: Referral asks sent and referral pipeline started.
Ship the season pass
- Add saved organization profile and signatory block.
- Ship the $399 monthly pass with a 5-letter entitlement counter.
- Offer the pass to any buyer who mentions a second docket.
Deliverable: Season pass purchasable in production.
Second sector list
- Build 25 prospects in a second sector with a live docket.
- Send 25 messages with sector-specific provision examples.
- Refresh the message template with the best-performing opening line.
Deliverable: 25 sends in a new sector.
Throughput without quality loss
- Batch-process all queued letters and review each ledger for mismatches.
- Manually audit five random citations per letter before sending.
- Note every mismatch as a prompt or validation bug.
Deliverable: All queued letters delivered with audit notes.
Trust angle versus urgency angle
- Split remaining sends between the citation-audit angle and the deadline angle.
- Track reply rate by angle over 40 messages.
- Keep the winner as the default template.
Deliverable: Reply-rate comparison across two message angles.
Cut cost per letter
- Cache extraction and quote pools per docket.
- Measure tokens and dollars per delivered letter.
- Trim prompt size where quality holds.
Deliverable: Measured cost per letter under $6.
Week three truth
- Count paid, delivered, refunded and lost.
- Read every refund and non-reply reason.
- Cut the lowest-yield outreach step entirely.
Deliverable: Written week-three review with one cut decided.
Week 4 · Scale what works · 7 days
Referral and repeat sources
- Send 25 messages to referral-sourced and repeat-docket prospects.
- Post one public sample ledger in a professional community.
- Follow up on every warm thread older than four days.
Deliverable: 25 sends and one public sample posted.
Deadline crunch
- Deliver all letters with a docket close date inside seven days first.
- Send status updates to buyers waiting on regeneration.
- Confirm every delivered file opens correctly.
Deliverable: Zero overdue letters.
Ledger as the trust surface
- Polish the in-letter footnote table and CSV column names.
- Add a one-line audit instruction at the top of the ledger.
- Re-render two old letters to confirm nothing regressed.
Deliverable: Final ledger format shipped.
Fresh dockets
- Harvest newly opened dockets and add five named prospects per docket.
- Send 25 messages against the new dockets.
- Prioritize dockets closing before December 16.
Deliverable: 25 sends against newly opened dockets.
Convert singles to pass
- Offer the season pass to every single-letter buyer with a second docket.
- Add a two-line comparison of single versus pass on the pricing page.
- Track pass attach rate.
Deliverable: Pass offered to all eligible buyers.
Clear the queue
- Deliver remaining letters and regenerations.
- Send a delivery confirmation with the ledger attached.
- Ask for one referral per delivered letter.
Deliverable: Queue cleared and referral asks sent.
List every docket closing before December 16
- Build the final list of open dockets with deadlines inside the contest window.
- Rank prospects by deadline proximity.
- Prepare 48-hour reminder messages.
Deliverable: Ranked close-out list with reminder copy.
Days 29-30 · Consolidate · 2 days
Convert before the clock
- Send every 48-hour reminder with the exact remaining hours.
- Send checkout links to all unclosed warm threads.
- Deliver same-day on anything that pays.
Deliverable: All warm threads closed or explicitly lost.
Verified numbers only
- Reconcile Stripe payouts against the TrustMRR-connected account.
- Log gross collected, refunds, fees and ad spend for the contest window.
- Write a one-page post-mortem with the next 30-day plan.
Deliverable: Reconciled profit log and post-mortem.
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
- Conservative: 3 paid letters at the $249 single-letter price from roughly 150 outreach messages.
- Base: 10 paid letters at the $249 price, one refund, from roughly 300 outreach messages.
- Upside: 18 single letters plus 4 season passes, blended ARPU $276 across 22 customers.
- Payment fees modeled at a 4% blended rate (Stripe 2.9% + $0.30 per transaction, plus billing and dispute tooling).
- Ad spend is $0 in all scenarios; acquisition is manual one-to-one outreach, so paid marketing is not used.
- Operating cost is $6 per delivered letter in model and hosting cost plus roughly $25-40 of fixed monthly infra (Vercel, Supabase, domain, email).
- No revenue is collected before launch day 6, and all payments route through the TrustMRR-connected account.
- Customers are counted as paid orders inside the November 16 - December 16, 2026 window only.
- All figures are projections from stated assumptions, not verified revenue or existing customers.
Conservative
3 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $30
- Paid marketing
- $0
- Hackathon profit
- $717
- Operating costs
- $45
- Business profit
- $672
Base case
10 customers- Refunds
- $249
- Chargebacks
- $0
- Payment fees
- $100
- Paid marketing
- $0
- Hackathon profit
- $2,141
- Operating costs
- $105
- Business profit
- $2,036
Optimistic
22 customers- Refunds
- $498
- Chargebacks
- $0
- Payment fees
- $243
- Paid marketing
- $0
- Hackathon profit
- $5,331
- Operating costs
- $190
- Business profit
- $5,141
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $2,490
- Payment fees (3.9%)
- $97
- Paid marketing
- $0
- Hackathon profit
- $2,393
- Business profit
- $2,288
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.
All scenarios are projections built from stated assumptions. They are not verified revenue, not existing customers and not a forecast of results.
Main risks and validation experiments
Risk 1The model cites a provision incorrectly despite quote-locking, producing a letter the buyer cannot file.
Server-side exact string validation of every quote against source text; the drafting step may emit quote IDs only, and any ID that fails validation is dropped from the letter.
Run ten real proposed rules through the pipeline and manually audit five citations per letter for page and section accuracy.
If more than one in ten audited citations is wrong after validation, stop selling until the extraction and validation path is rebuilt.
Risk 2Extraction fails on scanned or multi-column PDFs, blocking the core workflow on live dockets.
OCR fallback, low-confidence detection, and a manual text-paste intake path that still preserves page breaks.
Test extraction on ten real proposed rules from five agencies and record failure rate by document type.
If more than 30% of tested dockets cannot be processed end to end, narrow to agencies whose documents parse cleanly before taking money.
Risk 3Not enough federal dockets open in the 30-day window, so prospects have no live deadline to act on.
Monitor open comment periods every Monday and Thursday and keep three sectors in rotation; prioritize dockets closing before December 16.
Count open proposed rules across three sectors on day 2 and again on day 14; track prospects per live docket.
If fewer than 15 usable open dockets exist at any weekly check, add a fourth and fifth sector immediately or shift to a longer-cycle offer.
Risk 4Professional buyers refuse to file anything AI-drafted regardless of the citation ledger.
Lead with the audit surface, not the model: the buyer reviews every citation in minutes and rewrites the prose freely.
In 20 outreach messages, ask directly whether an audit-ready draft is acceptable; log objection rate and wording.
If more than 70% of replies reject AI drafting outright, reposition to a citation-ledger-only tool for human-written letters.
Risk 5Concierge free-draft delivery does not scale past a few dozen prospects and eats the build days.
Cap free previews at eight per week and only for dockets closing within 21 days; generate them through the product, not by hand, once day 6 passes.
Track minutes spent per free preview and conversion from preview to paid over the first 15 previews.
If a preview takes over 45 minutes and converts under 10%, replace the offer with a paid $49 issue-section-only product.
Risk 6One-time pricing anchors low and blocks the monthly pass later.
Show the $399 season pass on the same pricing page from day 6 and offer it to every buyer with a second docket.
Offer the pass to all buyers from day 16 and track attach rate.
If zero pass sales occur after 15 single-letter buyers, retire the pass and add a per-docket rush fee instead.
Hour-by-hour: the first 24 hours
Hours 1-2
- Create the Next.js repo, Vercel project and Supabase schema for dockets, segments, quotes, letters and orders.
- Register docketdraft.com and point DNS at Vercel.
Output: Deployed skeleton with a working health route and committed database schema.
Hours 3-4
- Download three real proposed-rule PDFs from three agencies.
- Build pdfjs extraction that preserves page and character offsets, and test on all three.
Output: Extraction output with verified page markers on three real dockets.
Hours 5-6
- Write the section-boundary prompt and the quote-extraction prompt.
- Run both on one docket and manually check ten returned quotes against the source text.
Output: Quote pool JSON with a written accuracy note on the manual audit.
Hours 7-8
- Draft the fixed letter structure spec (addressee, executive summary, 3-5 issues, requested changes, closing).
- Write the drafting prompt constrained to quote IDs only.
Output: Letter spec file committed plus a working drafting prompt.
Hours 9-12
- Generate one complete sample letter and citation ledger from a live docket.
- Manually audit every citation in that sample and fix any mismatch.
- Publish the sample as the first public demo artifact.
Output: One audited demo letter plus CSV ledger that can be sent to prospects on day 7.
Full 30-day launch report
Read the full generated report
DocketDraft - 30-Day Launch Report
1. Product opportunity and positioning
DocketDraft sells one artifact: a submission-ready public comment letter plus a citation ledger that pins every argument to a quoted line, its page and its section in a proposed federal rule. The buyer is a solo public-affairs consultant or the single policy person inside a 1-10 person advocacy nonprofit, coalition or state association. The trigger is not a budget cycle - it is a docket opening with 10-35 days left on the clock.
That trigger is the commercial argument. The buyer is already committed to filing, the deadline creates urgency without any education budget, and the deliverable either exists or does not. There is no procurement, no integration and no regulatory approval between the founder and the first payment.
The alternative set is narrow and bad: a chat assistant writes fluent prose with invented section numbers, which is a liability in a document an agency will read; outside counsel costs more than the engagement earns; doing it by hand costs a week the buyer does not have.
JEV scored this concept 64.3/100 (band C), strongest on speed to first revenue (8.19) and reachable customers (7.8), weakest on monetization (4.87) and operational simplicity (4.9). This plan attacks those two weaknesses directly: the repeat-purchase season pass ships in week three rather than being retrofitted later, and the pipeline stays stateless on managed services with no human review step and no stored client data beyond the uploaded rule.
2. Smallest sellable MVP
The smallest sellable thing is a single paid flow: paste a docket URL or upload a proposed rule, answer five stance questions, see a watermarked preview of issue section 1, pay $249, receive a DOCX letter and a CSV citation ledger by email. Everything else is deferred. The preview matters more than the feature list - it is the proof that makes a deadline-pressured buyer pay before the letter is finished.
Explicitly not built: monitoring, alerts, state legislation, hearings, filing into regulations.gov, collaboration, seats, and any stance recommendation. Each excluded item adds operational surface without adding a dollar to the first 30 days.
3. Technical implementation plan
One Next.js app on Vercel, Supabase Postgres, pdfjs extraction that preserves page offsets, a long-context model for segmentation and drafting, Stripe Checkout, and Resend for delivery. The critical design choice is the quote lock: the drafting call receives a pool of verbatim quotes with IDs and may emit only those IDs, never free text citations. The server re-validates every returned quote against the source string and drops mismatches before rendering. That single constraint is the difference between a defensible letter and a liability.
Build runs days 1-5, launch day 6. The hardest risk is extraction on scanned and multi-column Federal Register PDFs; if page offsets drift, every citation is wrong, so OCR fallback and a manual text-paste path land on day 13.
4. Pricing and monetization
$249 per docket letter, one-time, including one regeneration pass. $399 per month for a season pass covering up to five letters, offered from day 16. Model and hosting cost per letter is measured on day 20 and must hold under $6; at $249 the letter is never sold below cost. The pass exists because the scored weakness is monetization: a consultant filing five comments a quarter should not pay five times, and a one-time price alone never compounds.
5. First-customer acquisition
No ads. The channel is one-to-one outreach to named commenters and solo consultants found in public regulations.gov comment threads, plus association staff directories and LinkedIn. Every message names a specific open docket and its close date and offers one free drafted issue section within 24 hours. The free section is generated by the product, capped at eight per week, and only for dockets closing within 21 days. Follow-ups go out 48 hours before each deadline with the remaining hours stated. Every buyer is asked for one referral to another organization that commented on the same docket.
6. 30-day execution roadmap
Days 1-5: build. Extraction, segmentation, quote lock, letter composition, export, Stripe and delivery. Day 6: live payment and public demo. Day 7: 60 named prospects, first ten messages.
Days 8-14: outreach volume and the first sale. Target day 11. Days 12-14 fix extraction failures, add OCR fallback, and reconcile the midpoint numbers against the base scenario.
Days 15-21: repeat revenue and cost control. Season pass ships day 16, a second sector list goes out day 17, message angles are split on day 19, cost per letter is measured on day 20, and one underperforming outreach step is cut on day 21.
Days 22-28: referral sourcing, deadline crunch delivery, ledger polish, and a ranked close-out list of every docket closing inside the contest window.
Days 29-30: final 48-hour reminders, then reconciliation of Stripe payouts against the TrustMRR-connected account and a written post-mortem.
7. Revenue and profit scenarios
All three scenarios are projections from stated assumptions, not verified revenue.
Conservative: 3 letters at $249, $747 collected, no refunds, $30 in processing fees, $0 ad spend, $45 operating cost.
Base: 10 letters at $249, $2,490 collected, one $249 refund, $100 in processing fees, $0 ad spend, $105 operating cost.
Upside: 22 customers blending 18 single letters and 4 season passes at $276 average, $6,072 collected, $498 refunded, $243 in processing fees, $0 ad spend, $190 operating cost.
Hackathon profit deducts refunds, chargebacks, payment fees and paid marketing only; hosting, tooling and model delivery costs are excluded from ranking but subtracted in business profit. Payment fees assume a 4% blended Stripe rate. Ad spend is zero because acquisition is manual.
8. Main risks and validation experiments
Citation error is the killer risk. Ten real rules are audited by hand; if more than one in ten audited citations is wrong after validation, selling stops.
Extraction failure on scanned PDFs. Ten dockets from five agencies are tested on day 2; over 30% end-to-end failure means narrowing to agencies that parse cleanly.
Too few open dockets. Open comment periods are counted every Monday and Thursday; below 15 usable dockets triggers adding sectors the same week.
Buyer refusal of AI drafting. Twenty messages ask directly; over 70% rejection forces a reposition to a citation-ledger-only tool.
Concierge previews eating the calendar. Minutes per preview and conversion are tracked across the first fifteen; over 45 minutes at under 10% conversion replaces the offer with a $49 paid issue-section product.
9. First 24-hour action plan
Hours 1-2: repo, Vercel, Supabase schema, domain. Hours 3-4: extraction with page offsets tested on three real rules. Hours 5-6: segmentation and quote-extraction prompts with a ten-citation manual audit. Hours 7-8: letter structure spec and the quote-ID-constrained drafting prompt. Hours 9-12: one complete sample letter plus ledger, fully audited, published as the demo artifact for day 7 outreach.
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 →Bridge Brief
Turns a founder's own LinkedIn connections export into a ranked warm-intro map against a specific target list, with a ready-to-send ask for every path.
AnswerProof Audit Kit
A white-label audit generator that lets small SEO and content agencies produce a client-ready PDF showing how ChatGPT, Perplexity, and Gemini describe a client's brand versus its competitors, complete with a 30-day fix list.
Zéro Écart
A fixed-price catch-up service for French small businesses that are months behind on bookkeeping: upload a bank export plus a pile of receipts, and receive a finished, importable accounting journal, renamed receipts, and an itemised list of exactly what could not be matched.