LLMs have become standard tools in developer workflows, accelerating everything from boilerplate generation to architectural review.
- Code completion and generation (Copilot, Cursor)
- Automated code review and refactoring
- Test generation from function signatures
- Bug explanation and auto-fix suggestions
- Documentation and docstring generation
- SQL, regex, and shell command synthesis
Clinical text is dense, structured, and high-stakes — LLMs assist with understanding and generation while humans remain in the loop.
- Summarising long patient records for clinicians
- Differential diagnosis support from symptoms
- Translating medical jargon for patients
- Drug interaction lookup and triage assistance
- Clinical trial protocol drafting
- Medical imaging report generation
Personalised tutoring at scale — LLMs can adapt explanations to a student's level, generate practice problems, and give instant feedback.
- Adaptive tutoring in math, science, language
- Essay feedback and revision suggestions
- Concept explanation at multiple difficulty levels
- Quiz and flashcard generation from notes
- Language learning conversation partners
- Lecture summarisation and Q&A
From first draft to polished copy, LLMs augment creative work — not replacing the human voice but accelerating production and exploration.
- Copywriting and marketing content at scale
- Story ideation and narrative outlining
- Script and screenplay development
- SEO article drafting and rephrasing
- Social media and email campaign writing
- Poetry, lyrics, and experimental prose
Knowledge work inside organisations — from support tickets to strategy documents — is being compressed and automated with LLM-powered tools.
- Customer support chatbots with knowledge bases
- Contract analysis and legal document review
- Earnings call and report summarisation
- HR onboarding and policy Q&A systems
- Meeting transcription and action-item extraction
- Internal knowledge management and search
LLMs are becoming research accelerators — digesting literature, hypothesising, and generating experimental designs at speeds no human team can match.
- Systematic literature review over thousands of papers
- Hypothesis generation from experimental data
- Protein structure and drug molecule design (AlphaFold paradigm)
- Research paper writing and editing assistance
- Code generation for data analysis pipelines
- Cross-disciplinary knowledge synthesis
The most powerful applications combine LLM reasoning with tools, memory, and external systems — building autonomous agents that plan, act, and iterate without constant human input.
- Web browsing, form-filling, and data scraping agents
- Autonomous software engineering (SWE-agent, Devin patterns)
- Multi-agent pipelines: planner → executor → reviewer
- Computer-use agents that operate GUIs directly
- Email and calendar management assistants
- Financial data analysis and report generation workflows
- CI/CD pipeline monitoring and auto-remediation
- Research agents that search, read papers, and synthesise findings