Corpus vs Unanet ProposalAI: AI for SOQs at engineering firms
By Yermek Ibray, Founder & CEO, SalemWise Solutions. Every statement about Unanet ProposalAI comes from Unanet ProposalAI's own public pages as of October 10, 2026, linked inline and listed under Sources. Corpus statements can be checked in the live demo.
Who this is for
You run an engineering firm of 20 to 500 people, or you lead its business development. You want AI help on qualification-based RFQs and SOQs, and Unanet ProposalAI keeps coming up, often because the firm already uses Unanet. This page compares what each product is built to do, and how to test both on one real pursuit.
What each product is built to do
Unanet ProposalAI
Unanet says ProposalAI helps marketing, business development and proposal teams move from a complex RFP or RFQ to a stronger first draft faster (as of October 10, 2026). It analyzes solicitation requirements, organizes response obligations and searches approved firm content. That content is approved source material such as past proposals, project sheets, resumes, firm overviews and project descriptions. Teams can upload documents in batches, connect supported sources such as SharePoint, or use documents tied to an opportunity in Unanet CRM. Unanet says ProposalAI can be evaluated as a standalone solution, and that the CRM connection makes the workflow more fluid. It describes ProposalAI as best suited for longer, qualification-based RFP and RFQ responses, producing editable draft content that many teams finish in Word or InDesign. Under "See the Win Before You Pursue", it says you can instantly assess opportunities with data-backed insights. It runs in a secure, private cloud environment that Unanet says complies with federal cybersecurity requirements. Unanet does not publish ProposalAI prices on that page.
Corpus by SalemWise
Corpus by SalemWise is page-cited AI search over a company's own documents. Engineering firms are the first worked example. For them, Corpus reads the project record itself: reports, drawings, specs, calc packages and bid tabs. Each answer cites the document and page. Click a citation and the source page opens with the sentence highlighted. If the archive does not hold the answer, Corpus is designed to say "Nothing on file." On a new RFQ, it suggests bid/no-go scores with each scored factor linked to its sources. Fit, relationship and capacity stay your call. It drafts SOQ sections from your past projects with page citations, and flags what the archive cannot support for a person to write. It builds the fee with the math done by code. Arithmetic done in code is only as good as its inputs and formulas, so a qualified person still checks them. Set up for you: Corpus is cloud-agnostic. It runs on AWS, Azure, Google Cloud, or your own servers, with an open-weight model and no public AI API.
Side by side
| Unanet ProposalAI | Corpus by SalemWise | |
|---|---|---|
| Built for | Marketing, BD and proposal teams at AEC firms | Principals, BD leads and engineers at engineering firms |
| Main job | A first draft of a proposal from the solicitation and approved firm content | Find facts in the project record and draft from them, with page citations |
| Works from | Past proposals, project sheets, resumes, firm overviews, project descriptions | Project files: reports, drawings, specs, calc packages, bid tabs |
| Ties to other systems | Unanet CRM optional; SharePoint where supported | No CRM needed; the pilot uses a copy of your project PDFs |
| Go/No-Go | Opportunity assessment ("See the Win Before You Pursue"); how it scores is not described on the page we read | Suggested scores, each scored factor linked to its sources |
| Fee | Unanet says pricing stays with your team | Fee build with the math done by code, checked by a person |
| Where it runs | Secure, private cloud | Set up for you: AWS, Azure, Google Cloud, or your own servers |
| Price | Not published; book a demo | $5,000 pilot; setup from $21,000, then $1,500 to $2,000 a month, no per-user fees |
When to choose Unanet ProposalAI
- Your firm already runs Unanet CRM, and opportunity data lives there.
- You have a good library of approved proposal content: past proposals, project sheets, resumes and descriptions.
- Your main bottleneck is turning a long solicitation into an organized first draft.
- A private cloud service fits your IT policy.
If your approved content is current and complete, ProposalAI drafts from it, and Corpus would add less.
When Corpus fits
- The approved content runs out. The fact the RFQ asks for is in a geotech report or a bid tab, not in any past proposal.
- You need each claim in the SOQ cited to a page a reviewer can open.
- You want a suggested bid/no-go score and a fee build from the same past projects.
- You would rather hear "Nothing on file" than get a confident guess.
- You need the files to stay on your own servers.
The two can work together. Corpus pulls cited facts from the record. Your team adds them to the content library ProposalAI drafts from.
How to test both on one real RFQ
Do not decide from a feature list. Run one pursuit through both and count what you could check.
- Pick one RFQ you are pursuing now. Use your own files, not a sample.
- Write down five facts the SOQ must prove. For example: three similar projects from the last five years, the test standard you specified on one of them, and the unit prices from a recent bid tab.
- Ask each tool for those five facts. For each answer, open the source and time how long it takes to confirm it.
- Ask one question you know your files cannot answer. Note which tool says so and which one guesses.
- Compare the drafts your team would actually submit, and the hours it took to get there.
For Unanet ProposalAI, ask for a demo on your own content. For Corpus, start with the live demo. It holds a fictional sample firm's archive, built from real public engineering documents: about 7,000 pages of plan sets, geotech reports, specs, bid tabs and public RFQs. Then run your real RFQ in the pilot.
FAQ
Is Corpus a Unanet ProposalAI alternative?
For some firms. ProposalAI drafts from approved proposal content. Corpus drafts from the project record and cites the page. If your proposal library is strong, ProposalAI may be enough. If the facts live in project files, Corpus covers that gap.
Do we need Unanet to use Corpus?
No. Corpus does not need a CRM or an ERP. The pilot runs on a copy of your project PDFs with no integrations.
Where do our files go with Corpus?
Set up for you: Corpus is cloud-agnostic. It runs on AWS, Azure, Google Cloud, or your own servers. It uses an open-weight model and nothing goes to a public AI API. In the pilot, your files are never used to train or fine-tune a model. In full deployment, fine-tuning on your files happens only if you agree in writing. We don't have a SOC 2 report, which is one reason you can choose your own hardware.
Next step
Book a free 30-minute call. Bring one RFQ you are pursuing now and tell us where the facts for it live. We will tell you plainly whether it is a Corpus problem, a Unanet ProposalAI problem, or both.
If it fits, the next step is the $5,000 pilot on one real pursuit. It takes about 2 to 3 weeks. We load a copy of up to 3 years of your project PDFs and run your RFQ through Corpus: a go/no-go scorecard, an SOQ draft with page citations to check, and a fee build. There is no integration work. The $5,000 is credited toward setup if you sign a deployment within 60 days of the pilot's end. Full deployment is from $21,000 setup, then $1,500 to $2,000 a month, with no per-user fees.
Want to look first? Try the live demo or read more about Corpus. Comparing other tools? See all our side-by-side guides.
Sources
Checked October 10, 2026:
- Unanet ProposalAI for AEC: what it does, the content it uses, standalone evaluation, best-suited proposals, Word and InDesign, opportunity assessment, private cloud
Unanet and ProposalAI are names of their owner. SalemWise Solutions is not affiliated with Unanet. Check Unanet's pages for its current offer.
