Shrishan AI is the AI system that powers STCWBOOKER, an AI-driven maritime training ecosystem for seafarers, training institutes, and employers. It is not a separate product — it is the intelligence layer connecting course authoring, assessment, oral viva examination, knowledge search, and analytics across the platform.
Every capability below reflects what is actually built into STCWBOOKER today. Capabilities marked In development are not yet live — we don't claim features that don't exist yet.
A no-login chat assistant on the STCWBOOKER homepage that answers questions about courses, fees, batches, institutes, and general STCW/DG Shipping/SOLAS topics using STCWBOOKER's own data and an approved maritime knowledge library.
A tiered search engine that answers maritime and platform questions by checking, in order: fixed platform intents, STCWBOOKER's internal knowledge base, course/training-material content, and regulatory circulars (STCW, SOLAS, MARPOL, DG Shipping, IMO). Open, unrestricted internet search is intentionally excluded by design.
Converts uploaded source documents (PDF/Word) into structured courses: modules, lessons, learning objectives, and draft assessment questions, for institute staff to review and publish.
A spoken oral-examination module: AI-generated questions and branching scenarios drawn from course content, real-time spoken Q&A, AI scoring and first-person feedback on each answer, with results feeding into the formal assessment record alongside faculty-graded viva.
Orchestrates written assessments, AI viva, and faculty viva as stages of one competency framework, with automatic difficulty rating of modules/topics based on batch performance and configurable pass thresholds.
Generates multiple-choice, true/false, multiple-response, sequence, drag-and-drop, matching, and hotspot-image questions from a topic or lesson at a chosen difficulty level, for faculty review before use.
Scores spoken viva answers and generates personalised, first-person feedback per answer. For written assessments, AI is used only to turn computed item-analysis numbers into a plain-language narrative — it does not invent or alter the underlying scores.
Identifies a student's weakest topics from assessment history and generates a narrative highlighting revision priorities; in the viva module this extends to a personalised study-path recommendation matched to specific courses.
Item-analysis and coverage dashboards for assessments and the knowledge base: question difficulty, batch performance trends, and knowledge-base coverage/gap reporting.
Scheduled crawling and ingestion of trusted regulatory sources into the knowledge base, plus admin tooling for reviewing and approving AI-sourced content before it reaches students.
A dedicated regulatory knowledge tier covering STCW, SOLAS, MARPOL, IMO, and DG Shipping circulars, used both by the public assistant and by institute-side compliance tooling.
The knowledge base currently uses a local text-matching method as a placeholder for retrieval; a full semantic embedding search is planned but not yet live.
A trusted-domain web search tier (restricted to an approved allowlist, not open internet search) is built but not yet enabled pending live search API credentials.
The knowledge assistant checks sources in order of reliability: STCWBOOKER's own fixed answers first, then its internal knowledge base, then course and training-material content, then regulatory circulars (STCW, SOLAS, MARPOL, IMO, DG Shipping). Open, unrestricted internet search is intentionally excluded — answers are grounded in STCWBOOKER's own approved content.
Shrishan AI is the intelligence layer behind STCWBOOKER. It powers course authoring, oral viva examination, assessment question generation, evaluation and feedback, analytics, and maritime knowledge search across the platform.
STCWBOOKER uses AI across several areas: turning uploaded documents into structured courses, generating and scoring oral viva examinations, drafting assessment questions, analysing student performance, and answering maritime training questions through a tiered knowledge-search assistant.
Yes. The AI Course Authoring tool converts uploaded source documents into structured modules and lessons with draft learning objectives and assessment questions, which institute staff review before publishing.
Yes. STCWBOOKER can AI-generate multiple-choice, true/false, multiple-response, sequence, drag-and-drop, matching, and hotspot-image questions from course topics or lesson content, at a chosen difficulty level.
Yes. Shrishan AI Viva conducts spoken oral examinations with AI-generated questions and branching scenarios, scores each answer, and gives feedback in real time, alongside a separate faculty-graded viva option.
Yes, for oral viva answers, where AI scores responses and generates feedback. For written assessments, AI is used only to explain computed item-analysis results in plain language — it does not calculate the underlying scores.
Yes. STCWBOOKER identifies a student's weakest topics from their assessment history and can recommend specific courses to address those gaps.
The public maritime assistant and the approved Knowledge Library are available to any visitor without an account. Course authoring, AI viva, assessment tools, and analytics are available to logged-in students, faculty, and institute administrators.