Three tiers, month-to-month. The price you see is the price you pay. Free during the invite-only beta.
1 user. Independent consultants and one-person shops.
Best for: 1-3 bids per quarter.
Up to 5 users. Small firms running multiple bids per month.
Best for: consultancies and nonprofits with up to 5 staff.
Up to 15 users. Firms and BD teams with shared library.
Best for: BD teams running 50+ bids/year.
Free during the invite-only beta. No credit card required.
| Feature | Solo | Boutique | Team |
|---|---|---|---|
| Discovery & analysis | |||
| Daily opportunity feed (SAM.gov, Texas SmartBuy, Bonfire) | |||
| Solicitation analyzer (Section L/M extraction) | Unlimited | Unlimited | Unlimited |
| Compliance matrix + Excel export | |||
| Fit scoring against capability profile | |||
| Incumbent intelligence (USAspending) | |||
| Sources Sought / RFI response templates | |||
| Proposal generation | |||
| Drafts from your past wins | 10 / mo | Unlimited | Unlimited |
| Discovery call audio analysis | |||
| Capture pipeline (Kanban) | |||
| Workspace & access | |||
| Users | 1 | Up to 5 | Up to 15 |
| Role-based access | |||
| Tenant-isolated workspace | |||
| SSO (Google, Microsoft) | |||
| Audit log export | |||
| Support | |||
| Response time | 48 hrs email | 24 hrs email | Slack, same-day |
| Onboarding | Self-serve | Self-serve | Founder kickoff call |
Direct API integrations, not screen scrapes. We dedupe across sources, score against your capability profile, and surface the ones worth your time.
Federal solicitations, sources sought, pre-solicitations, RFIs. Daily ingest, NAICS pre-filtered to your profile.
Past contract awards by NAICS + agency. Powers the incumbent intelligence panel on every project.
State of Texas procurement. State-and-local opportunities in your geography that don't post to SAM.gov.
City, county, university, and K-12 district procurement portals. Curated subdomain coverage rather than full firehose.
Need a source we don't cover? Email sara@sparkdigitalinc.com — we add new ones based on customer pull.
We started with two cybersecurity tenants on day one. Every architecture decision had to hold up to a security review.
Every tenant runs in a workspace enforced at the database layer through Postgres Row-Level Security. Your library, capability profile, and proposals never cross tenant boundaries — even at the AI prompt level.
Every action against your data is logged: uploads, analyses, proposal generations, exports, role changes, captures. Owners and admins see the full audit log under Settings.
Documents live in Cloudflare R2 with server-side encryption. Database traffic over TLS. Anthropic and OpenAI bound by zero-retention agreements — your content is never used to train models.
Owner, admin, member, viewer roles per tenant. Viewers can review without editing; admins invite and audit. Single sign-on (Google, Microsoft) on the Team tier.
Drag the inputs. We'll show you what one win has to be worth for Proposal Engine to pay for itself.
Anywhere from $5K to $10M. Median for boutique consultants is $50K–$300K.
At Boutique pricing ($4,788/yr), you need a single win at $1,995 to break even.
Your projected revenue is 75× the annual cost.
Industry numbers as defaults. Your actual win rate climbs once you're scoring fit before you bid — historically the biggest time-sink for boutiques is bidding work that was never going to land.
Proposal Engine is in a free, invite-only private beta right now. If you have an invite code, you get full access to upload, analyze, and draft — no credit card, no trial clock. Don't have a code? Email sara@sparkdigitalinc.com and we'll get you in.
Anyone with a login. On Solo, that's just you. On Boutique you can invite up to 5 teammates. On Team you can invite up to 15. We don't charge per-seat pricing on top of the plan — the seat count is included.
Yes — strictly. Every tenant runs in an isolated workspace enforced at the database level (Postgres Row-Level Security). Your documents, drafts, and capability profile are never visible to other tenants, never used to train AI models, and encrypted at rest and in transit.
No. All plans are month-to-month. Cancel any time from your account settings. Annual prepay (with a discount) is on the roadmap but not required.
Upgrade or downgrade any time. Changes prorate to the day. If you hit Boutique's limits one month and drop back to Solo the next, you only pay Boutique pricing for the days you were on it.
Redeem your invite code and look around — that's the demo. Beta access gives you the full product, free. We deliberately don't gate it behind sales calls. If you have a specific question, email sara@sparkdigitalinc.com.
Not yet. Proposal Engine is built for unclassified procurement intelligence (FOUO at most). FedRAMP Moderate authorization is on the roadmap for the Team tier; reach out if it's a hard requirement.
Every opportunity is scored against your capability profile (NAICS codes, primary services, set-aside qualifications, geography). Strong matches surface to the top with explainable chips so you see why each row showed up. Once you capture an opportunity, a second-pass AI fit score (Haiku) gives you a Go / Lite-bid / No-bid recommendation in seconds.
Dismissed opportunities are tombstoned per-tenant. They disappear from your feed and pipeline, and re-ingesting the same opportunity from SAM.gov tomorrow won't bring it back. You can restore any dismissed item from the Dismissed list on the pipeline page.
The drafter pulls retrieval context from your library — past proposals, capability statements, methodology docs — and is explicitly prompted to never invent customers, certifications, or capabilities not present in your materials. Each section also lists which library docs fed it, so you can verify.
Yes. Every section save records who edited it, when, and whether the change came from manual editing, AI refinement, or initial generation. The pipeline card and section header show the most recent editor with their initials avatar; the per-section history panel surfaces the last 25 edits.
Database in Supabase (US-East). File storage in Cloudflare R2 (encrypted at rest). AI inference via Anthropic and OpenAI under zero-retention agreements (your content is not retained past the API call). Background jobs run on Inngest. All traffic over TLS.
Sign up, drop in a solicitation, and have a compliance matrix and fit score waiting for you 60 seconds later. No credit card. No demo call.