MansirMansir Muhammed
About·AI Layer

AI Layer

Research Intelligence Stack

A record of the models and tools that sit inside the research workflow and what each one is actually used for. It changes as the work changes.

Synthesis & Analysis
Claude
Anthropic · Claude Sonnet / Opus

Primary synthesis engine — literature analysis, research drafts, code architecture, and complex reasoning across all active projects

claude.ai
Code Generation
GitHub Copilot
GitHub / Microsoft · GPT-4o / Claude integration

In-editor code assistance across the full platform build — component generation, refactoring, and API design

github.com
Drafting & Ideation
ChatGPT
OpenAI · GPT-4o

Early-stage drafting, ideation, and structured reasoning for complex research problems and communication frameworks

chatgpt.com
Literature Review
Elicit
Elicit AI · Research-tuned LLMs

Citation extraction, systematic literature review, and deep research grounding across academic sources and primary documents

elicit.com
Reference Management
Zotero
Corporation for Digital Scholarship · Reference manager

Primary citation and reference library across all research projects — annotation, bibliography generation, and source tracking

zotero.org
Social Media Intelligence
Grok
xAI · Grok 3

Real-time social media data curation, event monitoring, and field-level signal extraction for active research interventions

grok.com

The pattern here is: model does the repetitive or computationally intensive part — surface a source, draft a synthesis, generate a code block — and a human decides whether the output holds up, what it means, and whether it goes anywhere. No finding published on this platform leaves AI as the last checkpoint.

This stack is updated as the workflow evolves. Lab AI tools — tools built in-house — are documented separately.