SARAP
Sharq University
Open Demo
Sharq University · Research Infrastructure

Research with Evidence.
Discover with Intelligence.

SARAP connects evidence discovery, research design, academic writing, verification, review and institutional completion in one traceable research environment. AI assistance remains subject to researcher verification and academic governance at every stage.

Open Demo Explore the Platform
SARAP research lifecycle
  1. 01
    Discover

    Literature discovery and source selection.

  2. 02
    Analyze

    Evidence extraction, structuring and research-gap analysis.

  3. 03
    Develop

    Research problem, objectives, questions, hypotheses and methodology.

  4. 04
    Write

    Evidence-grounded manuscript development with claims and citations.

  5. 05
    Review

    Research integrity, citation integrity and methodological consistency.

  6. 06
    Complete

    Supervisor-governed research completion and institutional record.

Stages advance on recorded researcher and supervisor decisions — never automatically.

Evidence-grounded research

Claims are linked to recorded evidence, not to generated text.

Researcher-verified AI

Every AI suggestion requires an explicit researcher decision.

Supervisor governance

Completion depends on a recorded supervisor decision.

End-to-end traceability

Source, evidence, claim, citation, manuscript and record stay connected.

Multilingual research

Uzbek, Russian and English sources within one research environment.

The research problem

Modern research workflows are fragmented, and traceability is the first casualty

Sources live in databases, references in a manager, drafts in documents, reasoning in notes, feedback in email, reporting in institutional systems. The chain from source to institutional record exists only informally.

Today · fragmented
  • Literature databases
  • Reference managers
  • Documents
  • General AI tools
  • Methodology notes
  • Citation tools
  • Supervisor feedback
  • Institutional reporting

Each tool holds part of the research record. The connections between them exist only in the researcher's memory.

With SARAP · connected
  1. 01Source
  2. 02Evidence
  3. 03Claim
  4. 04Manuscript
  5. 05Review
  6. 06Institutional record

One traceable chain, from the source that was screened to the institutional record that closes the project.

The chain SARAP keeps intact
SourceEvidenceClaimManuscriptReviewInstitutional record
Platform at a glance

Six connected research stages

Each stage produces an artefact the next stage consumes. Stages are advanced by recorded researcher decisions and supervisor governance — never automatically by AI.

01

Discover

Literature discovery and source selection.

Search strategy, multilingual results, screening decisions and a recorded decision log for every included or excluded source.

02

Analyze

Evidence extraction, structuring and research-gap analysis.

Evidence records with field-level provenance, evidence strength, themes and research gaps traceable back to the evidence they came from.

03

Develop

Research problem, objectives, questions, hypotheses and methodology.

A traceable design chain from evidence and gaps through to design, sampling and analysis plan, with suggested, draft and verified states.

04

Write

Evidence-grounded manuscript development with claims and citations.

Outline, drafting and a claim registry where each claim carries its supporting evidence, citation state and verification state.

05

Review

Research integrity, citation integrity and methodological consistency.

Inspection of the manuscript against evidence, claims, research questions and design — with findings recorded, not silently corrected.

06

Complete

Supervisor-governed research completion and institutional record.

A completion gate derived from recorded states, a supervisor final decision, and a structured institutional research record.

Evidence intelligence

Evidence support and citation presence are separate dimensions

SARAP records evidence as structured objects with field-level provenance. A claim can carry a complete citation and still be unsupported by the recorded evidence — SARAP shows both states rather than merging them.

Evidence grounding

Claims point to recorded evidence, not to generated prose.

Claim–citation alignment

Citation completeness and evidence support are assessed apart.

Contradictory evidence

Opposing evidence is surfaced as a finding for the researcher.

Insufficient evidence

Thin support is recorded rather than quietly accepted.

Source verification

Inclusion and exclusion decisions carry a recorded reason.

Metadata integrity

Incomplete reference metadata is reported, never invented.

Evidence architecture
  1. Source
  2. Evidence record
  3. Claim
  4. Citation
  5. Manuscript

A citation records where a statement came from. Evidence support records whether the recorded evidence actually sustains it. SARAP keeps these two dimensions separate.

Research design intelligence

Design objects developed as a chain, with the decision left to the researcher

Evidence and gaps inform the research problem, objectives, questions, hypotheses and method. SARAP compares alternatives; it does not select the methodology.

  • Research problem
  • Research objective
  • Research questions
  • Hypotheses
  • Conceptual framework
  • Variables / constructs
  • Research design
  • Unit of analysis
  • Population
  • Sampling strategy
  • Data sources
  • Analysis plan
Governance principle

AI suggestion ≠ researcher decision

Suggested, draft and verified are distinct recorded states. Nothing becomes a final research decision without an explicit researcher action.

Writing & research integrity

Evidence-grounded writing, inspected against the project's own record

Drafting works from a claim registry. Review then inspects the manuscript against evidence, claims, research questions and the recorded design, and logs findings for the researcher to resolve.

Writing chain
EvidenceClaimCitationDraftVerification

Claim support

Whether recorded evidence sustains the statement.

Citation completeness

Whether the reference is present and complete.

Unsupported statements

Statements with no evidence record behind them.

Contradictory evidence

Recorded evidence pointing the other way.

Terminology consistency

Constructs used consistently across sections.

Methodological consistency

Claims consistent with the recorded design.

Research-question alignment

Whether sections answer the recorded questions.

Recorded findings

Findings are logged, never silently auto-corrected.

Human-in-the-loop governance

SARAP assists. The researcher decides. The supervisor governs.

The platform records which of the three acted at each point, so academic responsibility always rests with a named person.

Assists

SARAP AI

Suggests, structures, compares and flags. It never records a research decision on its own.

  • Suggests structure
  • Compares alternatives
  • Flags inconsistencies
  • Explains findings
Decides

Researcher

Evaluates each suggestion, edits it, accepts or rejects it, and carries responsibility for the record.

  • Evaluates suggestions
  • Edits content
  • Accepts or rejects
  • Verifies records
Governs

Supervisor

Reviews academic quality at checkpoints and records the final decision on research completion.

  • Reviews checkpoints
  • Requests revision
  • Records approval
  • Governs completion
  • AI suggestion ≠ researcher decision
  • Researcher verification ≠ supervisor approval
  • Supervisor approval ≠ external peer review
  • SARAP completion ≠ publication acceptance
For researchers

Structure the work, keep the reasoning

  • Structured research workflowOne environment for the six stages instead of scattered tools and files.
  • Evidence traceabilityEvery claim can be traced back to the evidence record and source behind it.
  • Reduced fragmentationScreening decisions, evidence, design and drafting stay in one project context.
  • Clearer research designProblem, questions, hypotheses and method are developed as a connected chain.
  • Evidence-grounded writingDrafting works from a claim registry linked to recorded evidence.
  • Integrity checksUnsupported claims, contradictory evidence and citation gaps are surfaced as findings.
  • Transparent verificationEach object shows whether it is suggested, drafted or verified.
  • Supervisor-ready recordsCheckpoints present the recorded state rather than an informal summary.
Open Demo
For universities

Institutional research infrastructure

  • Standardised workflowsThe same lifecycle vocabulary across faculties and doctoral programmes.
  • Transparent supervisionCheckpoints show what was reviewed, by whom and with what decision.
  • Evidence and citation traceabilityClaims, citations and evidence remain inspectable after the fact.
  • Institutional research recordsCompleted projects close into a structured, archivable record.
  • Governance checkpointsStage transitions depend on recorded conditions, not on navigation.
  • Research quality visibilityAttention is directed to specific recorded exceptions, not to a score.
  • Doctoral research managementPortfolio-level view of stage, blockers and verification state.
Explore Institutional Research
Platform preview

Representative views from the research workspace

Structural previews of the current platform. Each corresponds to functionality available in the demo environment.

SARAP · Research Workspace
Research Workspace

Project context, lifecycle position, readiness and blockers in one persistent frame.

SARAP · Evidence Matrix
Evidence Matrix

Structured evidence records with field-level provenance and verification state.

SARAP · Research Design
Research Design

The traceable chain from evidence and gaps to design, sampling and analysis plan.

SARAP · Claims & Citations
Claims & Citations

A claim registry showing evidence support and citation completeness separately.

SARAP · Research Review
Research Review

Findings against evidence, claims, research questions and methodology.

SARAP · Institutional Record
Institutional Record

Completion gate, supervisor decision and the archived research record.

Open Demo
About SARAP

Sharq AI Research Assistant Platform

SARAP is developed as part of Sharq University's research infrastructure initiative, around responsible AI-assisted academic research, evidence integrity, research governance and institutional research management. It is currently a proof-of-concept environment presented for academic review; all research content in the demo is illustrative.

Evidence before generation

Research writing starts from screened sources and structured evidence records — not from unsourced text.

Researcher verification before authority

AI output carries no academic authority until a named researcher evaluates, edits and verifies it.

Supervisor governance before completion

A project cannot be recorded as complete without a supervisor decision on academic quality.

Traceability throughout the lifecycle

Every recorded object keeps its origin, its decision history and its current verification state.

Build research on evidence.
Keep every decision traceable.

Explore how SARAP connects research evidence, academic reasoning, verification and institutional governance across the research lifecycle.

Open Demo Learn More