Evidra
Sign in
Enterprise Response Intelligence

Every answer.
Backed by evidence.

Turn your company's existing knowledge into verified responses for RFPs, security questionnaires, DDQs and enterprise requests.

Start your response librarySee how it works

Free for 14 days on the Team plan.

Knowledge to evidence to verified response
Built for
The product in one frame

One answered question, with its evidence

A real answer from Evidra's own security assessment: the passages it was drafted from with document names and page numbers, a confidence signal, the reviewer's edit and comment, and the trail of who did what.

EvidraVendor security assessment, question 2.1, approved
2.1Data protectionHigh confidence80%Approved

Is customer data encrypted in transit and at rest?

Format: Yes or no with detail
Answer

Yes. Connections between browsers and the platform, and between the platform and each managed service, use TLS. Data at rest, including the database, the passages and the files, is encrypted by Google Cloud with provider managed keys.

Show the original draft

Yes. Connections between browsers and our platform, and between the platform and each managed service, use TLS for encryption in transit. Data at rest is encrypted by Google Cloud for the database, passages, and files, and by our storage providers.

Evidence2 sources
  • Evidra security overviewpage 1relevance 78%
    ... Encryption Connections between browsers and the platform, and between the platform and each managed service, use TLS. Data at rest is encrypted by Google Cloud for the database, the passages and the files. Provider managed keys are used throughout; customer managed keys are not offered today. ...
  • Evidra privacy policypage 2relevance 67%
    ... Where data is processed Authentication, the application database and uploaded files are held on Google Cloud in the European Union multi region. Embeddings, question extraction and drafting run on Vertex AI on Google Cloud in the Netherlands region. ...
Activity trail
  1. Evidra teamdrafted this answer from the knowledge library: 2 sources, high confidence
  2. Evidra teamedited the answer
  3. Evidra teamcommented
  4. Evidra teamapproved the answer
  5. Evidra teamsaved the approved answer to the answer library
How it works

From your documents to a verified response

Five steps, run as a background job with progress per question. The model only ever sees passages from your own library.

Step 01 of 05

Company knowledge

Past responses, policies, product documentation and compliance records, uploaded once.

Each document carries a category, an owner, an effective date and an optional expiry date, so an answer drafted from an expired policy is visible as such.

Step 02 of 05

Ingestion

Each document is read page by page and split into passages that keep their page numbers.

Each page is read and split into passages that keep their page numbers, so evidence can point back to the exact place in the document.

Step 03 of 05

Retrieval

Every question is matched against your passages only, never against the open web.

Hybrid vector and keyword search over one index per organization. No query from another organization can reach it.

Step 04 of 05

Evidence

The passages that support the answer are attached with document name and page.

The drafting prompt receives only retrieved passages and must cite them. No citation, no answer: the question is marked No evidence found.

Step 05 of 05

Verified response

A reviewer confirms the answer, an approver signs it off, and the export carries the evidence.

Approved answers join the answer library with their evidence, so the next questionnaire reuses them first.

How our AI works

An AI pipeline that has to show its sources

The same stages run on every questionnaire. The models read and write, but a rule in code decides what reaches your reviewers: no citation, no answer.

  1. 01

    Reads every page

    Policies, past responses and product documents are read page by page and split into passages that keep their page numbers, so every answer can point to the exact page it came from.

  2. 02

    Finds meaning, not just keywords

    Google text embeddings and keyword ranking work together in a hybrid search over one isolated index per organization. A question finds the right passage even when it uses different words.

  3. 03

    Reads the questionnaire for you

    Spreadsheets are parsed for sections, numbering, yes or no columns and character limits. Gemini reads Word and PDF questionnaires and returns each question with its section and answer format.

  4. 04

    Drafts only from your evidence

    Gemini writes each answer from the retrieved passages alone, in your organization's tone and inside the question's format and character limit, and cites the passages it used.

  5. 05

    Checks every citation in code

    A draft that cites nothing it retrieved is rejected before anyone sees it. The question is marked No evidence found, with a note on which document would answer it.

  6. 06

    Scores confidence and reuses approved answers

    Each answer carries a confidence signal from retrieval strength and citation coverage. Approved answers are matched to new questions by meaning and reused while their sources stay current.

The models are Gemini and Google text embeddings on Vertex AI, in a European Union region. Customer documents are never used to train them. How your data is handled

Integrity

When there is no evidence, it says so.

The platform never invents an answer. If no passage in your library supports a response, the question is marked No evidence found and the platform suggests which kind of document would answer it.

This is what makes the AI claim honest. A drafted answer can only contain what a cited passage contains, so a reviewer can check every sentence against the source in seconds, and the absence of an answer is a signal your team can act on rather than a guess a buyer will find later.

EvidraVendor security assessment, question 4.2
4.2AssuranceNo evidenceNo evidence found

Do you carry cyber liability insurance? Provide the coverage limit.

Format: Yes or no with detail
Answer

No evidence found in your knowledge library.

Nothing was drafted, because no source document supports an answer. A certificate of insurance or a summary of insurance coverage would answer this question.

No evidence attached.
Verification workflow

Verified by people who own the answer

Drafting is the start. Every answer moves through the people responsible for it, and the product records each step.

  • Evidence linked
  • Reviewer roles
  • Activity trail
  • Answer library

Contributor

Uploads documents, corrects extracted questions and comments on drafts. Sees every answer and its evidence, changes none of the verification state.

Reviewer

Edits drafts, keeps or removes evidence passages, flags answers that need another owner and approves answers that have evidence attached.

Approver

Signs off the questionnaire, exports it, and is the only role that can approve an answer without evidence, which the activity trail records as an override.

EvidraVendor security assessment, question 1.1, edited and approved
1.1Access controlHigh confidence75%Approved

How do users sign in to the platform?

Format: Free text
Answer

Users sign in with an email address and password, which must be verified, or with a Google account. Firebase Authentication on Google Cloud handles both sign-in methods.

Show the original draft

Users sign in to the platform using an email address and password, which requires verification, or by using a Google account. We collect your name, email address, and password hash, or the identifier provided by Google when you sign in with a Google account. Firebase Authentication on Google Cloud handles these sign-in methods.

Evidence3 sources
  • Evidra security overviewpage 2relevance 75%
    ... Users sign in with an email address and password, which must be verified before a workspace opens, or with a Google account. Sessions are server side cookies that expire after 14 days. Each user belongs to one or more organizations with one of three roles: Contributor, Reviewer or Approver. ...
  • Evidra privacy policypage 1relevance 51%
    ... We collect only what the service needs to work. Account data: your name, email address and password hash, or the identifier Google gives us when you sign in with a Google account. Organization data: the organization name, industry, questionnaire types, tone setting, members, roles and invitations. ...
  • Evidra security overviewpage 1relevance 48%
    ... - Firebase Authentication on Google Cloud for sign in: email and password with verified addresses, and Google accounts. - Firestore on Google Cloud for organizations, members, documents, passages, questionnaires, answers, evidence, comments, usage and the activity trail. The browser ...
Activity trail
  1. Evidra teamdrafted this answer from the knowledge library: 3 sources, high confidence
  2. Evidra teamedited the answer
  3. Evidra teamapproved the answer
  4. Evidra teamsaved the approved answer to the answer library

Editing an answer keeps its evidence link: the reviewer chooses which passages still support the new text. Removing all evidence marks the answer Unverified, and it stays that way until an Approver overrides it, which the activity trail records.

Security and data handling

Practices, stated plainly

How customer documents are isolated, encrypted, kept out of model training and deleted, written as practices you can check.

  • Isolated index per organization

    Your passages live in their own search index. No query from another organization can reach them.

  • No training on customer data

    Documents, answers and questionnaires are used to answer your questions and for nothing else.

  • Deletion propagates within minutes

    Deleting a document removes its chunks from the index and its file from storage within minutes.

  • Encryption in transit and at rest

    TLS on every connection. Storage providers encrypt files, rows and index data at rest.

  • Data held in the European Union

    Files and records are held on Google Cloud in the European Union multi region, and the models run on Vertex AI in a European Union region.

Read the security page
Security overviewIsolation, encryption, deletion and the no training rule, written as practices you can check.
Security page
For teams and enterprises

Built for the teams that answer, and the teams that check

Presales, security and bid teams work in one organization with named seats and explicit roles. Larger organizations get audit export and priority processing.

Talk to sales

Seats

Named seats per plan, with invitations by email and role changes recorded in the activity trail.

Roles

Contributor, Reviewer and Approver on every organization, enforced on the server, not only in the interface.

Audit export

Every action on an answer, document or member is exportable as a dated record for your own audit files.

Founder

Who is building this

Evidra is built in Abuja, Nigeria, around one rule: every answer must trace back to a passage in your own documents.

Emmanuel (Michael) Ugbaja

Emmanuel (Michael) Ugbaja

Founder

Every answer Evidra drafts cites the passages it came from, and the pipeline rejects any draft that cannot. That rule is enforced in code, not suggested in a prompt.

LinkedIn
Pricing

Plans for every size of team

Four tiers in USD, from a single team to a large organization. Prices are per organization per month.

Starter

For one team answering a handful of questionnaires a month.

$149per month
Seats
3
Pages indexed
2,000
Talk to sales

Team

Trial available

For presales and security teams with a steady questionnaire load.

$499per month
Seats
10
Pages indexed
20,000
Start a 14 day trial

Business

For larger organizations with big document libraries.

$1,499per month
Seats
30
Pages indexed
100,000
Talk to sales

Enterprise

Custom seats, pages and terms for large organizations.

Custom
Seats
Custom
Pages indexed
Custom
Talk to sales

Priced in USD for every region.

See full pricing and the comparison table

Start your response library

Upload the documents your team already trusts. Every answer that follows is drafted from them, cited and reviewed before it leaves.

Start your response libraryTalk to sales