Agent Orchestration Literature Review

Automated Synthesis and Validation
of Academic Literature

Ditenagai oleh sistem Agentic AI. Berbeda dengan asisten AI biasa, di AutoCite seluruh agen AI bekerja sama secara kolaboratif untuk menelusuri, memvalidasi, dan menyusun literatur akademik dengan akurasi tinggi.

Deterministic Information Flow

Multi-Agent Orchestration Architecture

AutoCite distributes operational workloads across bounded, isolated sub-routines to systematically govern target verification and pipeline efficiency.

MODULE 01

Topic Decomposition

Evaluates multi-word prompt arrays to identify semantic core properties, generating normalized boolean variables suitable for algorithmic database ingestion.

MODULE 02

Bibliometric Retrieval

Interfaces directly with the OpenAlex open-access repository. Applies structural exclusion constraints based on publication data completeness and recency.

MODULE 03

Provenance Verification

Evaluates incoming dataset properties to strictly ensure structural compliance with standard academic citation frameworks (APA 7th formatting rules).

MODULE 04

Cross-Doc Integration

Synthesizes information vectors across confirmed abstracts, mapping divergent assertions or methodological contradictions objectively.

Functional Capabilities

System Framework Specifications

The pipeline utilizes standardized scientific infrastructure routines to facilitate data discovery and mitigate collection bias.

OpenAlex Registry Integration

Executes algorithmic queries directly against open bibliographic records, mitigating typical manual discovery discovery latency and index capture bias.

Syntax Standardization

Deterministic normalization rules check all generated tokens against exact APA 7th layout paradigms, stabilizing referencing outputs.

Heuristic Filtering Rules

Applies explicit pruning thresholds: non-abstract items are categorically eliminated from consideration to safeguard structural completeness.

Objective Contradiction Analysis

Designed to map heterogeneous empirical outcomes. The aggregation layer identifies conflicting conclusions within literature clusters without subjective curation bias.

orchestration_engine.py

$ python orchestration_engine.py --run

[LOG] Executing Topic Decomposition...

[LOG] Fetching OpenAlex Remote Entities...

→ Compiling 3 semantic text vectors

→ Validating syntactic compliance standards

Undergraduate Thesis Project

Research and Development Personnel

AutoCite is developed within an undergraduate research setting to analyze structural automated documentation engineering frameworks.

Willy Wijaya
Lead Researcher

Willy Wijaya

Undergraduate Student
Informatics Department
Academic Advisor I

H. Hengky Anra, S.T., M.Kom.

NIP. 197503251999031005
Software EngineeringInstitutional Profile
Anggi Perwitasari
Academic Advisor II

Anggi Perwitasari, S.T, M.T.

NIP. 198908192019032012
Software EngineerInstitutional Profile