Blog · Page 26
Field notes.
Page 26 of 31. Browse the archive of RFP workflows, grounded-AI architecture, and proposal operations notes.
The proposal week bracketing July 4th
Federal buyers don't move. Commercial buyers don't move. Reviewers do. A short field note on the bid weeks bracketing the July 4th holiday and the four things to triage now.
Shipped: the inline verify button in drafts
Hover any drafted sentence in the proposal builder and a verify button surfaces the source block, the entailment trace, and the timestamp of the last KB update. Shipped this week.
How we evaluate retrieval quality on our own corpus
Our gold set, the metrics we track, the eval harness on a laptop, the regression-guard CI job, and the directional numbers we'll publicly stand behind. Long read.
The RFP kickoff meeting that saves two weeks
A 60-minute kickoff that decides ownership, win themes, the compliance scaffold, and the review calendar. Run badly, it costs you two weeks. Run right, it pays for itself by Friday.
Month two: the questions we got from readers
Two months into writing this blog daily. Six questions readers sent, paraphrased. What I learned from the questions, and where the blog is going next.
DDQ Anatomy, Part 4 of 4: operations and vendor management
The closing section of a vendor DDQ. Incident response from the operational side, business continuity, vendor risk management, and the questions that decide whether you're a vendor procurement will renew.
Anatomy of a 28-page California state RFP
A walkthrough of a representative 28-page state-of-California RFP — the structure, the scoring framework, the compliance language, and the patterns that recur across California state procurements.
A compliance matrix in thirty minutes, not two days
A compliance matrix is the scaffold the rest of the proposal is built on. Most teams build it by hand from a PDF, and it takes two days. Here's the path from PDF to matrix in 30 minutes.
The Friday 4pm review is the wrong ritual
End-of-week color-team reviews are the default in most proposal shops. The reviewers are tired, the writers can't fix anything, and the meeting becomes a tracker entry instead of a review. There's a better cadence.
The hallucination budget, per claim
Treat hallucination as a cost: each claim in a draft has a probability of being mis-attributed. Here's how we budget it, how we trade latency against grounding strength, and why the budget is per-claim, not per-draft.
Why we priced in public on day one
Public pricing in enterprise RFP software is a posture signal, not a conversion tactic. A founder's note on why we put numbers on the site before we had a sales team.
Inside the ingest pipeline: parse, extract, index
How a PDF becomes searchable KB blocks. LlamaParse for parsing, structural-plus-semantic extraction, pgvector indexing with HNSW. Where each stage wins and where it falls over.
Prefer to see the product?
Take the 5-minute tour, or start a trial workspace and see PursuitAgent draft answers with citations.