Review Mockup

AI Signer Assessment Flow

This corrected mockup shows the intended product direction: the app itself asks questions in ASL through a saved signer avatar, watches the covenant child or adult sign back, translates the response, and builds a doctrine-aware assessment for communion, baptism, or membership.

Scenario

A pastor opens the communion assessment with a covenant child who knows ASL. The app plays stored signer clips, reads the child's signed responses, and returns an assessment tied to RPCNA doctrine.

Workflow

AI-Led Interview Journey

The point of this flow is to let the communicant answer directly through sign language, while the church receives a structured assessment grounded in confession, catechism, and testimony.

1

Open a Prepared Assessment Pack

The pastor selects a communion, baptism, or membership pack that already contains stored ASL question videos, alternate clarifications, and the doctrinal rubric to apply.

2

Play Cached ASL Signer Clips

The app presents each question through an AI-generated signer avatar clip that has already been rendered and saved, so the same approved prompt can be reused every time.

3

Capture the Communicant's Signed Response

A camera records the covenant child or adult signing back, while OpenAI vision tracks hand shape, movement, body language, and pace to form a translation candidate.

4

Map the Response to Doctrine

The translated answer is grouped under Christ, sin, faith, repentance, sacrament understanding, and vow readiness so the system can assess what was actually communicated.

5

Produce an Assessment for Session

The app returns a recommendation with linked evidence clips, translated responses, and follow-up flags so the church can decide whether admission, baptism, or membership is appropriate.

Screen One

Launch the AI Assessment

The first screen is where the pastor chooses the prepared interview pack and confirms the communication mode before the app begins asking questions on its own.

Supported Communicant AssessmentAI signer pack selected
Candidate Profile

Samuel M.

  • Primary communication: ASL
  • Assessment mode: communion admission
  • Pastor supervising: Rev. James K.
  • Guardian present: mother
Assessment Pack

Lord's Supper Readiness

  • 18 saved signer clips
  • 4 stored clarification clips
  • 3 sacrament-specific follow-up clips
  • 1 doctrine rubric bound to the pack
Signer Library

Human-like ASL avatar prompts have already been rendered and approved, so the app reuses them instead of generating a new question video during the interview.

Standards in Scope
  • Westminster Confession of Faith chapter 28 on baptism and chapter 29 on the Lord's Supper
  • Westminster Larger Catechism on faith, repentance, sacraments, and self-examination
  • Westminster Shorter Catechism on God, sin, Christ, salvation, church membership, and the sacraments
  • RPCNA testimony, forms, and membership-vow context for a credible profession

Screen Two

Avatar Asks and Vision Interprets

The live screen is the center of the product: the signer avatar asks a question, the communicant signs back, and the app translates and scores the answer in real time.

Question 12 of 18Communion meaning and worthy use
Signer Avatar Stage
ASL question clipWhy do Christians take the bread and cup?
Play saved clipReplayUse clarification clip
Candidate Camera Feed
Live response captureCommunicant signing to camera
Frame burstRecord response
Vision Translation Stream
Prompt clipASL signer avatar asks: WHY CHRISTIAN TAKE BREAD AND CUP?
Vision glossCandidate signs: JESUS DIE FOR ME. BREAD-CUP SHOW BODY BLOOD. BELONG JESUS PEOPLE.
Plain-language translationThe candidate appears to say that communion belongs to Jesus, remembers his death, and is for those who belong to him.
Confidence + flagConfidence 0.86. Follow-up suggested on self-examination before eating and drinking.
Doctrine Sync
  • Christ's death connected to the Supper
  • Believer participation appears understood
  • Self-trust rejected in favor of faith in Christ
  • Self-examination still needs one direct follow-up

Screen Three

Assessment Output for Session

Once the exchange is complete, the app should return a doctrine-based conclusion with the linked evidence that supports it.

AI Assessment SummaryCommunion recommendation with evidence
Recommendation

Likely ready for communion, with one follow-up on self-examination

The system judges that the candidate communicated the core Gospel and sacramental meaning with enough clarity for Session review, while flagging one final question on worthy receiving.

Assessment Matrix
Christ understoodResponses consistently identify Jesus as Savior, Son of God, crucified, and risen.
Sin and forgiveness understoodThe candidate communicates wrongdoing, sorrow, and the need for Christ's forgiveness.
Faith and repentance expressedSigned responses show trust in Jesus and desire to follow him with God's help.
Sacrament meaning understoodThe bread and cup are connected to Christ's death, but one follow-up on worthy participation is still recommended.
Membership-vow readinessThe candidate appears able to affirm the substance of profession and submission, with support for signed communication.
Evidence Vault
  • 18 prompt clips linked to their exact doctrinal category
  • 12 response clips stored with translation snapshots
  • 4 low-confidence moments marked for human review
  • Export ready for Session minutes and pastoral follow-up

Component Library

What the Product Now Needs to Prove

These are the components the app needs if it is going to function as an AI assessment system rather than a manual note-taking interface.

Component

Assessment Pack Selector

Lets the church choose a communion, baptism, or membership workflow with a saved ASL clip library and the exact doctrinal rubric attached to it.

Component

Signer Avatar Playback Panel

Shows the AI signer asking the question in ASL, replaying the same approved clip, or switching to a stored clarification version without inventing a new prompt live.

Component

Candidate Response Capture

Records the covenant child or adult signing back so the system can preserve the response as evidence and re-run interpretation if needed.

Component

Vision Translation Stream

Turns signed and gestural communication into gloss, plain-language translation, confidence, and follow-up warnings for unclear or incomplete doctrinal answers.

Component

Doctrine Mapping Engine

Scores each translated response against RPCNA, confession, catechism, sacrament, and vow categories instead of leaving the result as a vague accessibility summary.

Component

Session Decision Workspace

Presents the recommendation, linked clips, and supporting rationale so elders can review what the AI saw before taking church action.