One rule for this page: a skill is only as strong as the work linked next to it. "No shipped project yet" means exactly that. Click a column header to sort.
How levels work
Learning - studied it or built small exercises; no shipped work yet.
Working - used it in a shipped project or paid work; can build with it on my own.
Strong - used it across several shipped projects or paid work; can design, debug and defend it in an interview.
The full table
59 skills. Proof links point at my projects page.
| Skill | Category | Level | Where I learned it | Used at | Proof |
|---|---|---|---|---|---|
| Python | AI & LLM engineering | Strong | Self-study, daily use for years | FoCDoT; contract evaluation; independent | MailU, Revenue Intelligence Engine, SecAudit, CopyPasta |
| LLM application development | AI & LLM engineering | Strong | FoCDoT work, then my own products | FoCDoT; independent | RevenueRoles, AGENYRA, MailU |
| Prompt engineering & output validation | AI & LLM engineering | Strong | Contract evaluation programs; building CopyPasta | Contract evaluation; independent | CopyPasta |
| LLM evaluation | AI & LLM engineering | Strong | Contract evaluation programs (senior reviewer, then lead) | Contract evaluation | Coding-eval archive |
| Evaluation task & dataset design | AI & LLM engineering | Strong | Contract evaluation programs | Contract evaluation | ~186 tasks with gold solutions + tests |
| RLHF & preference-data work (review and grading, not model training) | AI & LLM engineering | Working | Contract evaluation programs | Contract evaluation | - |
| RAG pipeline design | AI & LLM engineering | Strong | FoCDoT work and my own tools | FoCDoT; independent | MailU, adaptive-rag (extended) |
| AI agents (LangGraph) | AI & LLM engineering | Working | Extending adaptive-rag | Independent | adaptive-rag (extended) |
| Local LLM serving (Ollama) | AI & LLM engineering | Strong | Building local-first tools | Independent | CopyPasta, MailU |
| Vector search (pgvector, ChromaDB, Qdrant) | AI & LLM engineering | Strong | Building own products | Independent | RevenueRoles, MailU |
| Hybrid search & classification | AI & LLM engineering | Working | Building MailU and RevenueRoles | Independent | MailU, RevenueRoles |
| AI system design | AI & LLM engineering | Working | Designing own products | Independent | AGENYRA, RevenueRoles |
| Fine-tuning (LoRA/QLoRA) | AI & LLM engineering | Learning | Coursework and self-study | - | no shipped project yet |
| AI agents (CrewAI) | AI & LLM engineering | Learning | Self-study | - | no shipped project yet |
| MCP server development | AI & LLM engineering | Learning | Self-study | - | no shipped project yet |
| AI observability | AI & LLM engineering | Learning | Self-study | - | no shipped project yet |
| PyTorch | AI & LLM engineering | Learning | Coursework and self-study | - | older learning projects |
| AI audit / consulting | AI & LLM engineering | Learning | Self-study | - | no shipped work yet |
| TypeScript / JavaScript | Full stack | Strong | Building own products | Independent | SpiderWeb, AGENYRA, Relationship Intelligence Engine |
| React / Next.js | Full stack | Strong | Building own products | Independent | AGENYRA, Revenue Intelligence Engine, Shiksha-Mitra, RevenueRoles |
| HTML / CSS | Full stack | Strong | Self-study; every site I shipped | FoCDoT; independent | All live sites |
| Node.js | Full stack | Working | Building APIs | Independent | Relationship Intelligence Engine, SpiderWeb |
| FastAPI | Full stack | Strong | FoCDoT work and my own tools | FoCDoT; independent | RevenueRoles, MailU, CopyPasta, Revenue Intelligence Engine |
| Streamlit | Full stack | Working | Building MailU | Independent | MailU |
| Testing (unittest, pytest, Vitest) | Full stack | Strong | Evaluation work and projects | Contract evaluation; independent | Coding-eval archive, Revenue Intelligence Engine |
| SQL / Postgres | Data & pipelines | Strong | FoCDoT work and own products | FoCDoT; independent | SpiderWeb, AGENYRA, Relationship Intelligence Engine |
| DBMS | Data & pipelines | Working | College and projects | Independent | - |
| Entity resolution & deduplication | Data & pipelines | Strong | Building pipelines | Independent | Relationship Intelligence Engine, RevenueRoles |
| Provenance-first data modeling | Data & pipelines | Working | Building the Relationship Intelligence Engine | Independent | Relationship Intelligence Engine |
| Untrusted-archive ingestion | Data & pipelines | Working | Building SpiderWeb | Independent | SpiderWeb |
| Pandas / NumPy | Data & pipelines | Working | College and learning projects | - | older analysis projects (learning) |
| Excel | Data & pipelines | Working | College and ops work | - | - |
| Tableau | Data & pipelines | Learning | Self-study | - | no shipped project yet |
| Statistics | Data & pipelines | Learning | Coursework | - | - |
| AWS (EC2, S3, Lambda) | Cloud & DevOps | Working | FoCDoT work | FoCDoT | - |
| Docker | Cloud & DevOps | Working | Containerizing projects | Independent | adaptive-rag (extended) |
| Git / GitHub | Cloud & DevOps | Strong | Daily across all repos | FoCDoT; independent | github.com/ayushrai-hub |
| CI/CD (GitHub Actions) | Cloud & DevOps | Working | SecAudit workflows | Independent | SecAudit |
| Vercel deployment & custom domains | Cloud & DevOps | Strong | Shipping own sites | Independent | agenyra.space, revenueroles.in |
| Google Cloud | Cloud & DevOps | Learning | Self-study | - | no shipped project yet |
| Data structures & algorithms | Fundamentals | Working | College + evaluation task work | Contract evaluation | coding tasks: A*, Dijkstra, TSP |
| Operating systems | Fundamentals | Working | College + SecAudit host internals | Independent | SecAudit |
| Computer networks | Fundamentals | Learning | Coursework | - | - |
| System design | Fundamentals | Working | Designing own products | Independent | AGENYRA, SpiderWeb |
| Host security auditing / DFIR basics | Security | Working | Building SecAudit | Independent | SecAudit |
| SARIF reporting | Security | Working | Building SecAudit | Independent | SecAudit |
| Review leadership (senior reviewer, then lead) | Leadership & work style | Strong | Contract evaluation programs | Contract evaluation | - |
| Mentoring & peer review | Leadership & work style | Working | Review programs | Contract evaluation | - |
| Project management | Leadership & work style | Working | Running own products | Independent | RevenueRoles, AGENYRA |
| Technical writing | Communication | Strong | Writing a breakdown for every project | Independent | project breakdowns |
| Communication | Communication | Working | Work and review programs | FoCDoT; contract evaluation | - |
| VS Code | Tools & automation | Strong | Daily tool | All work | - |
| Google Colab | Tools & automation | Working | ML learning work | - | older learning projects |
| Gmail API & OAuth | Tools & automation | Working | Building MailU | Independent | MailU |
| Form abuse protection | Tools & automation | Working | Building the AGENYRA waitlist | Independent | AGENYRA |
| SEO / AI-search basics | Tools & automation | Working | RevenueRoles launch | Independent | RevenueRoles |
| n8n | Tools & automation | Learning | Self-study | - | no shipped project yet |
| Zapier | Tools & automation | Learning | Self-study | - | no shipped project yet |
| Salesforce (no-code) | Tools & automation | Learning | Trailblazer modules | - | no shipped project yet |
Contact
Email - ayushrai0211@gmail.com
Projects - project-showcase-ayushr.vercel.app
GitHub - github.com/ayushrai-hub
LinkedIn - linkedin.com/in/ayushrai02