Summary
Tech Assessment scores a technology you are evaluating and shows the reasoning behind every number and the research it drew on. You give it an invention disclosure, patent text, or a publication, either pasted in or uploaded. FirstIgnite researches the technology and scores it.
Each score comes from a separate evaluation, and each evaluation is scored on its own. That is why a technology can look strong commercially and still be held back by a weak patent position, instead of having that averaged away. Twelve evaluations are available in total, and the template you use selects which of them run and how much each one counts. For what each evaluation asks and how its scores are banded, see Evaluations: scope, limits and rubrics.
This article covers what an assessment is, how to create one, reading the result, controlling the scoring, what the assessment can and cannot find out on its own, and where to go next.
When to use
You need to decide whether a new disclosure justifies a patent filing budget.
You have more disclosures than time and need to rank them.
You are preparing for a licensing conversation, a diligence request, or an investor review, and you want a read you can defend.
You want to show an inventor where their technology looks strong and where the evidence is thin.
Before you begin
Find Tech Assessment in the sidebar under Mission Control.
Have your source material ready. You can paste text, upload files, or do both.
Supported upload formats are PDF, DOCX, DOC, PPTX, TXT, TEX, MD, and JSON, at 10MB or less per file.
Some parts of an assessment are available on Premium only. Every tier produces every evaluation in your template, the Overall Score, and the reasoning behind each number. For what changes between tiers, see Basic and Premium: what changes.
Step-by-step instructions
Create an assessment
Select Tech Assessment in the sidebar, then Create New Assessment.
The three-step build opens on Context. It shows the template it will use at the top, and you can change it there.
Paste your material into Technology Context, or add files under Upload Document.
The build lists attached documents underneath with their file size, and you can remove any of them. You need either text or one document to continue.
Select Next.
FirstIgnite checks that your input has enough substance to assess. If it does not, a message explains what is missing.
On Mode, choose how deep the research should go, then select Next.
Quality, around 20 minutes. Recommended, and the right choice for licensing, diligence, or anything leaving your institution.
Fast, around 10 minutes. Useful directional insight for internal screening and portfolio work, though less precise on very recent research.
Rapid, around 5 minutes. Immediate signal for exploration. Read it as a starting point rather than a conclusion.
On Details, enter an Assessment Title and an Internal Reference ID if you want them.
Both are optional. Leave the title blank and FirstIgnite generates one from your input.
Select Generate Assessment.
The assessment works through Initializing, Processing files, Researching, Evaluating, and Finalizing. You do not need to wait on the page. Select Go to Tech Assessments and come back later. You will receive an email once the report is finished.
The Context step of the build, showing the selected assessment template with its description, the Change Type control, the three build steps, and the Technology Context field above the upload area.
After the assessment finishes
Open the assessment from the Tech Assessments list. The result page leads with an Overall Score, then states what is holding the technology back, then breaks the score down by evaluation. Read it in that order rather than starting with the individual evaluations. See Read your assessment results.
Tips
Use Quality for anything leaving your institution. The faster depths are useful for triage, but they trade rigor for speed and can sound more certain than the evidence supports.
Upload documents instead of pasting summaries of them. The assessment reads the full text and picks up detail a summary drops.
Put everything you know into the input, including material that is not published anywhere. The assessment researches what is publicly findable, but unpublished performance data and disclosure history reach it only if you supply them. See Where the assessment's data comes from.
Enter an Internal Reference ID if you track disclosures in another system. It makes assessments easier to match up later.
Common issues
Next says you need to provide context. Add text in Technology Context, or upload at least one document.
A file was rejected. Files must be within the size limit and in a supported format. Rejected files appear in the attached list with the reason, so remove or convert them and upload again.
Next does not respond right after an upload. Uploads have to finish first. Wait for the spinner beside each file to turn into a document icon.
The assessment is still running. Research and scoring take time, and the page does not need to stay open. Return to the Tech Assessments list and reopen the assessment later.
Frequently asked questions
How long does an assessment take?
It depends on the research depth you choose. Quality takes around 20 minutes, Fast around 10, and Rapid around 5.
Do I have to stay on the page while it runs?
No. Select Go to Tech Assessments and reopen the assessment when it has finished.
Can I change the input after generating?
No. Each assessment records what was submitted, so the input is fixed once it runs. Create a new assessment with better input instead. You can rename an assessment at any time, and you can change how its scores are weighted after the fact.
Why do two assessments of the same technology score different things?
Three things can move the result. First, the template: it decides which of the twelve evaluations run and how much each one counts, so two templates are measuring different things. Second, the research: every assessment reads live public sources, and what is publicly findable changes over time, so a later run can find different evidence. Third, the reasoning: each evaluation weighs the evidence it finds rather than applying a fixed formula, so even the same input on the same template can come back with different scores. For the template, see Assessment templates: choosing, creating and editing. For how much movement to expect on a given score, see Evaluations: scope, limits and rubrics.
What do I do if a score looks wrong?
Start with what the result page says is holding the technology back, rather than the Overall Score on its own. See Read your assessment results.
Where to go next
Read your assessment results covers the result page in the order worth reading it: the Overall Score, then the plain-language posture, then the constraint that is actually limiting the technology, then the per-evaluation breakdown. Start here once your first assessment has finished.
Evaluations: scope, limits and rubrics explains what an evaluation is, lists all twelve, and gives the scoring bands. It also covers the limits worth knowing: why a re-run can score differently, what an unscored evaluation means, and how much confidence to place in a single number.
Assessment templates: choosing, creating and editing covers the template that decides which evaluations run and how much each counts, and how to build one that matches what your institution cares about.
Editing weights after an assessment is for changing the balance on a report that has already finished, and the one caveat that comes with it.
Where the assessment's data comes from explains the three kinds of data an assessment draws on, and which of them only you can supply. Read this if an assessment ever seems to have missed something.
Premium patent data: what it is and how to use it covers the patent search available on Premium, what it is good for, and where it stops. Read it before treating any result as a patent position.
Basic and Premium: what changes is a straight comparison of the tiers, covering what Premium adds before, during, and after a run.
Pre-flight checks: preview before you spend a run covers the Premium check that reads your input before you generate and tells you where it is thin, so you can improve it first.