Lifecycle
Research is not linear. Neither is this framework. 32 steps across 5 phases, AI-guided at every stage, human decision-making throughout.
From project genesis through manuscript development. Click any phase to jump to details.
From initial idea to search-ready. AI challenges assumptions, pressure-tests research questions, and validates alignment through iterative Socratic dialogue.
Fully autonomous pipeline. Multiple AI engines execute search, enrich metadata, screen for relevance and novelty threats, and produce deep reviews with no human intervention.
Researcher validates AI results, adjudicates edge cases, records evaluations, and conducts targeted gap-filling searches that loop back through the agentic pipeline.
Theory foundation through methods design. Each workflow follows Socratic planning, creation, AI enhancement, and critique. Includes Libby Diagram and full-project evaluations.
Every section follows a 5-step workflow: Socratic Planning, Outline, Draft, Critique, Enhance. AI guides, researcher approves. Iterative until publication-ready.
Phase I
From initial idea to search-ready. AI challenges assumptions, pressure-tests research questions, and validates alignment through iterative Socratic dialogue.
Study idea articulated. The initial spark where the researcher identifies a phenomenon, problem, or gap worth investigating.
Problem statement, motivation, overview, description, and rough draft abstract developed.
Iterative refinement of the project description narrative, sharpening the problem statement, motivation, and study overview.
Iterative process of articulating, testing, and refining the research questions that will drive the study.
Finalized research questions formally stated and aligned to the problem statement.
Structured evaluation of the project description and research questions together for alignment, clarity, feasibility, and methodological soundness.
Visual mapping of the research question relationships, construct interactions, and conceptual structure.
Design the search strategy: select databases, define inclusion/exclusion criteria, establish screening protocols.
Generate, refine, and finalize search terms with Boolean logic, synonyms, and domain-specific vocabulary.
Phase II
Fully autonomous pipeline. Multiple AI engines execute search, enrich metadata, screen for relevance and novelty threats, and produce deep reviews with no human intervention.
Steps 10-14 are orchestrated autonomously across multiple AI engines with automated scoring, classification, and threat assessment. No human intervention until Phase III. Results feed directly into all research outputs and reports.
Execute searches across selected databases, collect and deduplicate results.
Enhance article records with DOIs, abstracts, and bibliographic data via OpenAlex and CrossRef.
Retrieve missing abstracts using Abstract Snag, Grab AI, and other extraction tools.
Initial relevance and novelty threat screening of all articles against the project description and research questions using multiple AI engines.
In-depth AI-assisted analysis of flagged and promising articles.
Review notes and action items, relevance scoring with synthesis rationale, methodological and theoretical contribution assessment.
Threat level classification, key overlaps identified, key differences documented.
Detailed threat analysis narrative, key concerns identified, mitigation suggestions, final assessment and research positioning.
Phase III
Researcher validates AI results, adjudicates edge cases, records evaluations, and conducts targeted gap-filling searches that loop back through the agentic pipeline.
Multi-layered human review of AI results.
Confirm final article set ready for synthesis. Human review completed.
Targeted searches to fill gaps identified during the evaluation process. Researcher-directed queries for specific topics, methods, or theories that emerged as needs during review.
Selected articles from Step 17 loop back through Steps 11-14 via the Automated Agentic Pipeline. Results automatically become part of all Research Outputs & Reports.
Report 1
Each article assessed against the project description and research questions.
Report 2
Articles sorted by relevance score with comprehensive evaluation.
Report 3
Ranked listing of highest-impact articles identified in the review by citation count.
Report 4
Distribution of articles by journal, publication year, and related bibliometric metrics.
Report 5
Export-ready files for citation management.
Comments made by the researcher in the system are saved to the reports.
Phase IV
Theory foundation through methods design. Each workflow follows Socratic planning, creation, AI enhancement, and critique. Includes Libby Diagram and full-project evaluations.
AI-guided readiness assessment for theoretical approach, readiness scoring (7+/10 threshold), recommended theory count.
AI suggests verified theories with relevance scores. Seminal article verification via OpenAlex API. Theorist identification and seminal work references.
AI generates descriptions from seminal article context (abstract, citation, authors). Required before theory can be used in foundation.
AI synthesizes selected theories into cohesive narrative. Executive summary generation. Rich text editing for manual refinement.
AI validates foundation for consistency, coverage, and academic rigor. Identifies gaps, unresolved items, and theories not selected.
AI-guided readiness assessment for hypothesis formulation (7+/10 threshold). AI recommends number of hypotheses.
Type classification (Descriptive, Relational, Causal, Comparative). Theoretical grounding and expected effect size. AI enhance, validate, and create operations.
Visual mapping of construct interactions, hypothesis linkages, and conceptual structure.
AI-guided readiness assessment for construct definition (7+/10 threshold). AI recommends counts by role (Independent, Dependent, Mediator, Moderator, Control).
Role classification with multiple role flags. Impact on dependent variable documentation. Moderator/mediator commentary. AI enhance, validate, and create operations.
Per construct type (Independent, Dependent, Moderator, Mediator, Control). Variable definition with measurement approach. AI enhance, validate, and create operations.
AI-guided readiness assessment for methodology (7+/10 threshold). Coherence check against hypotheses, constructs, and variables.
7-section standardized document: Research Design, Sampling Strategy, Data Collection, Data Analysis, Validity & Reliability, Ethical Considerations, Measurement Challenges. AI create, enhance, validate, and critique per section.
Assessment of theoretical contribution strength, originality, and significance to the field.
Comprehensive evaluation of the full project for alignment, rigor, feasibility, and methodological soundness.
Phase V
Every section follows a 5-step workflow: Socratic Planning, Outline, Draft, Critique, Enhance. AI guides, researcher approves. Iterative until publication-ready.
Front Matter
Introduction
Literature & Theory
Methodology
Outcomes & Closing
Researcher establishes writing style, formatting standards, citation format, tone, and terminology conventions. These decisions guide all AI drafting.
AI asks targeted questions. Researcher answers and makes decisions about content, framing, and dependencies. Planning continues until the researcher is satisfied.
AI generates outline based on researcher decisions. Researcher reviews, revises, and approves the outline before any drafting begins.
AI drafts following the approved outline. Researcher reviews every paragraph, revises as needed, and decides when the draft is ready for critique.
AI provides expert peer review feedback. Researcher decides which feedback to accept, which to reject, and which to address differently.
AI suggests improvements in Track Changes format. Researcher accepts or rejects each change. Critique and Enhance repeat until the researcher decides the section is publication-ready.
Critique and Enhance repeat until the researcher decides it is publication-ready
Human-in-the-Loop
Human-in-the-Loop at every step: AI assists, researcher decides. No content advances without researcher approval. No changes are made without researcher consent. The researcher controls the pace, direction, and final output.
Comprehensive Manuscript Critique: Full manuscript review evaluating coherence across all sections, academic rigor, logical flow, citation integrity, and publication readiness. Can be run at any time.