Clear your inbound queue and know why each deal is in it
A review queue ranked by your own rubric, fit, evidence and readiness shown separately, and the same short interview for every founder you select.
- Tidewell FreightFreight audit software · Seed · raising USD 2.5M
- Thesis fit
- Strong
- Evidence
- Good
- Readiness
- Revenue data missing
- QuillstackDeveloper tooling · Seed · raising USD 3M
- Thesis fit
- Strong
- Evidence
- Thin
- Readiness
- Early
- Ferncastle HealthClinic scheduling · Seed · raising USD 1.8M
- Thesis fit
- Partial
- Evidence
- Good
- Readiness
- Ready for a call
Sample data. Companies, funds and people are made up.
More fitting decks than first meetings
Seed funds with steady inbound rarely lack deals. They lack the hours to read every deck closely and still explain, a month later, why one company got a meeting and another didn't.
- Your rubric orders the queue
Each application is checked against the thesis you published, so deals that fit your stage, sector and round size sit near the top with the reason written beside them.
- Three readings, never one score
Thesis fit, evidence quality and diligence readiness stay separate. A strong fit with thin evidence never looks the same as a weaker fit with every document in place.
- Passes you can explain later
A pass is a person's decision with a recorded reason. When the company applies again next year, the earlier deck, claims and reasoning are still on the record.
Every row says why it is there
The queue is ordered by your published rubric, but the order is never the whole story. Each deal shows its three readings and a short explanation with the criteria behind it, so an analyst can disagree with the order and see exactly where.
- Thesis fit: how the deal matches your hard rules and weighted preferences.
- Evidence quality: how much of what matters is backed by a cited source rather than a bare statement.
- Diligence readiness: what is still missing before the deal can move on.
- Tidewell FreightFreight audit software · Seed · raising USD 2.5M
- Thesis fit
- Strong
- Evidence
- Good
- Readiness
- Revenue data missing
AI suggestionWhy it is here: Stage, sector and round size met. Recurring revenue claimed on slide 7, not yet backed by a document.
- QuillstackDeveloper tooling · Seed · raising USD 3M
- Thesis fit
- Strong
- Evidence
- Thin
- Readiness
- Early
AI suggestionWhy it is here: Fits sector and stage. Most traction claims are founder reported, so evidence quality is low.
Unknown: Lead investor, asked as a question
- Ferncastle HealthClinic scheduling · Seed · raising USD 1.8M
- Thesis fit
- Partial
- Evidence
- Good
- Readiness
- Ready for a call
AI suggestionWhy it is here: Geography outside your rules, flagged as a mismatch for review. The exception for repeat founders may apply.
- BramblepayPayroll for contractors · Pre-seed · raising USD 900K
- Thesis fit
- Partial
- Evidence
- Thin
- Readiness
- Early
AI suggestionWhy it is here: Stage is earlier than your seed rubric allows. Worth checking against your pre-seed rubric.
Unknown: Round size, asked as a question
Sample data. Companies, funds and people are made up.
Your seed thesis, written as rules a reviewer can check
A sample seed rubric. Hard rules decide what is in scope, weighted preferences order what fits, exceptions name the cases you always want to see, and questions are asked whenever the deck is silent.
| Criterion | Type | Sample rule |
|---|---|---|
| Stage | Hard rule | Seed, or late pre-seed with recurring revenue |
| Geography | Hard rule | Company operates in the regions you invest in |
| Round size | Hard rule | Raising between USD 1M and USD 4M |
| Sector | Weighted preference | B2B software weighted 3, fintech infrastructure weighted 2 |
| Traction | Weighted preference | Recurring revenue, with how the founder defines it |
| Repeat founder | Exception | Review even when geography is outside the rules |
| Who is leading | Question | Asked whenever the deck doesn't say |
Rubrics are versioned. Publish a change and the next screen uses it; re-running an old deal creates a new run and keeps the earlier one. Funds with more than one strategy can keep a rubric per strategy or stage.
How a seed fund's week of inbound moves through Delsivo
From the application to the first partner meeting. Each step shows the release it arrives in.
- Pilot
Applications land on your portal
A public link for your website, private referral links for co-investors and portfolio founders, and records your analysts create for decks that arrive by email. Possible duplicates are flagged by company domain and founder, and an analyst confirms any merge.
- Pilot
Each deck is read with citations
Company facts and claims are extracted from the PDF, each linked to its slide. A chart that couldn't be read is flagged, not guessed.
- Pilot
Your seed rubric is applied
Hard rules, weighted preferences and exceptions, written by your team. Keep a separate rubric for opportunistic pre-seed cheques if your strategy has one.
- Pilot
The analyst queue fills in rubric order
Every row says why it is there: criteria met, criteria mismatched and criteria still unknown, each with its source.
- First release
Selected founders take a live AI interview
A short pitch with screen share, then an AI interviewer asks your questions plus follow-ups on the deck and its open claims, within the cap you set. Text is always an option.
- First release
Partners get a brief before the first meeting
Open claims, unknown criteria and timestamped interview excerpts, with suggested questions marked as AI suggestions.
Delsivo is at the design partner stage. Pilot features come first, first release features follow, and later features are on the roadmap. AI suggests and cites its sources; your team makes every decision.
Walk into the first meeting knowing what is still open
Before a partner meets a founder, Delsivo gathers what the record already knows and what it doesn't. Call preparation arrives in the first release.
- Open claimsEach with its label: Founder reported, Corroborated, Disputed or Unresolved, and where it came from.
- What is still unknownCriteria the deck and interview didn't settle, ready to ask in the meeting.
- What the founder saidInterview excerpts with timestamps, so a partner can play the moment instead of trusting a summary.
- Suggested questionsDrafted by AI from the open claims, labelled as suggestions, edited or dropped by the analyst.
A week for a seed fund analyst
Sample week. The interview and meeting brief are first release features.
- Monday
The weekend's applications are already read. The queue lists them in rubric order, each with the criteria it met, missed or couldn't confirm.
- Tuesday
You open the top six, invite three founders to the interview and pass on two with a short reason each. The partner approves the invitations before they go out.
- Wednesday
A deck claims USD 600K in annual recurring revenue. In the interview the founder says a third of it is one-off services. Both claims are kept, the claim is marked Disputed and a task to check it is assigned to you.
- Thursday
The partner reads the meeting brief for Tidewell Freight before the call: two open claims, one unknown criterion and three timestamped excerpts.
- Friday
After the meeting, a revenue schedule request goes to the founder through the portal. By the afternoon the upload is scanned, filed and waiting for your review.
What seed fund partners ask us
Design partner pilot, then Fund workspace
Lean seed funds with steady inbound are who the pilot is designed for: branded intake, deck analysis, your rubric, the analyst queue and the claim tracker. Interviews and meeting briefs arrive in the first release, in the Fund workspace. Prices are set after pilot usage is measured.
Bring your inbound to a design partner pilot
We start with a sample of your past decks to check extraction and your rubric, then run live intake with recommendations only your analysts see.