USEK Curriculum Watch

For the Provost

USEK's catalogue names 13 of 18 technology areas that leading universities teach. Three of its five gaps are among the areas peers added fastest since 2021.

Missing at USEK: AI agents, reasoning models, AI-assisted software work, quantum computing, green and sustainable computing. Since 2021, the 9 peers with archived catalogues went from 2 to 8 naming generative AI, from 5 to 9 naming quantum computing, and from 0 to 3 naming AI agents and AI-assisted software work. Quantum computing is now named by 13 of 16 peers. Like most peers, USEK does name cybersecurity, cloud, data, robotics, IoT, blockchain and AR/VR.

AI agents are named in 5 of 12 leading US catalogues and in the 2026-27 catalogues of LAU and USJ. LAU's catalogue lists courses titled Agentic AI and Large Language Models. Transformers and LLMs are named in 10 of 12. At USEK, GEIT215 is the only course description that names transformers. Four Computer Science courses need a refresh at course level.

A gap means the topic is missing from a catalogue description, which can be shorter than the syllabus. Each dean confirms against current syllabi before anything changes.

Decisions

What the Provost decides

Open items. None has a default.

Open

Send this slice to the three units below

Arts & Sciences, Engineering and General Education receive their course sheets with a request to confirm against syllabi.

Open

Set the date for deans to reply

The cascade table shows "Provost sets" until a date is chosen.

Open

Approve the watch cadence

The proposed cadence per source is in the Sources section.

Open

Choose the next slice

Business, engineering, health sciences, or another faculty.

Technology map

Where USEK stands across 18 technology areas

A filled cell means at least one course description in that institution's catalogue names the area. It does not measure depth. Red outline: an area USEK's 2025-26 catalogue does not name. Hover a cell for the count of course entries.

NamedNot namedUSEK gap

Institution	AI							Computing					Frontier technology
Institution	AI agents	Generative AI and LLMs	Reasoning models	Machine and deep learning	Computer vision	Responsible AI and tech ethics	AI-assisted software work	Cybersecurity	Cloud and DevOps	Data engineering and big data	GPUs and high-performance computing	Human-computer interaction and UX	Quantum computing	Robotics and autonomous systems	IoT, edge and embedded	Blockchain and Web3	AR, VR and spatial computing	Green and sustainable computing
USEK
Lebanese peers with course descriptions read
LAU
USJ
Balamand
LIU
Leading US universities
Carnegie Mellon
MIT
Stanford
Harvard
UPenn
Columbia
Cornell
Caltech
Illinois
Georgia Tech
UT Austin
Washington
Peers naming it, of 16	7	13	3	16	16	10	6	16	12	10	15	12	13	14	11	9	11	3

USJ, Balamand and LIU are read as whole documents; the others course by course. BAU, NDU, Haigazian and AUST publish titles only and are left out.

How it is evolving

What leading universities added between 2021 and 2026

The same catalogue pages read twice with the same rules: archived copies from 2021 (Internet Archive, April to November 2021) and live pages on 11 Nov 2026. 9 institutions: Carnegie Mellon, Cornell, Georgia Tech, LAU (Computer Science and Computer Engineering), MIT, Illinois, UPenn, UT Austin, Washington. Grey bar: 2021. Blue bar: 2026. USEK's earlier catalogues were not found, so USEK shows 2026 only.

Generative AI and LLMs

2 → 8USEK names it

Quantum computing

5 → 9USEK does not

AI agents

0 → 3USEK does not

Responsible AI and tech ethics

3 → 6USEK names it

AI-assisted software work

0 → 3USEK does not

Blockchain and Web3

3 → 6USEK names it

Cloud and DevOps

6 → 8USEK names it

Reasoning models

0 → 1USEK does not

Computer vision

8 → 9USEK names it

Data engineering and big data

5 → 6USEK names it

GPUs and high-performance computing

7 → 8USEK names it

Human-computer interaction and UX

7 → 8USEK names it

Green and sustainable computing

1 → 2USEK does not

Machine and deep learning

9 → 9USEK names it

Cybersecurity

9 → 9USEK names it

Robotics and autonomous systems

9 → 9USEK names it

IoT, edge and embedded

7 → 7USEK names it

AR, VR and spatial computing

7 → 7USEK names it

AI detail

AI topics, institution by institution

US rows count the courses whose official catalogue description names the topic; hover a number to see which courses. Lebanese rows show whether the documents read name it at all. n/c means only course titles were published, so no comparison is possible.

Institution	What was read	Transformers and LLMs	AI agents	Generative models	Reasoning models	Responsible AI	Deep learning	Reinforcement learning	Computer vision
USEK	2025-26 University Catalogue, all faculties	2	0	3	0	1	12	3	2
Lebanese peers
LAU	2026-27 catalogue: CSC, Applied AI, Data Science, Computer Engineering, Data Analytics (228 course entries, read in Chrome)	Named	Named	Named	Not named	Named	Named	Named	Named
USJ	2026-27 catalogue: Computer and Communications Engineering; Computer Science for Business	Named	Named	Named	Not named	Named	Named	Named	Named
Balamand	Arts and Sciences catalogues, undergraduate and graduate, updated Nov 2026	Not named	Not named	Not named	Not named	Not named	Named	Not named	Named
LIU	Computer Engineering and MS Computer and Communications Engineering descriptions (program PDFs)	Not named	Not named	Not named	Not named	Not named	Not named	Not named	Named
BAU	Computer Science program page: course titles only	n/c	n/c	n/c	n/c	n/c	n/c	n/c	n/c
NDU	BS Computer Science and AI minor pages: no course descriptions	n/c	n/c	n/c	n/c	n/c	n/c	n/c	n/c
Haigazian	Computer Science study plan 2026-28: titles only	n/c	n/c	n/c	n/c	n/c	n/c	n/c	n/c
AUST	BS Computer Science page: study plan, titles only (read in Chrome)	n/c	n/c	n/c	n/c	n/c	n/c	n/c	n/c
Leading US universities: number of course entries naming the topic
MIT	Official catalogue, 314 course entries	3	0	4	0	2	20	8	7
Stanford	Official catalogue, 114 course entries	4	1	6	1	7	6	1	2
Carnegie Mellon	Official catalogue, 526 course entries	29	4	23	4	9	52	13	28
Harvard	Official catalogue, 143 course entries	6	5	8	2	3	4	6	4
Columbia	Official catalogue, 81 course entries	5	2	0	0	0	3	2	2
Cornell	Official catalogue, 803 course entries	5	0	3	0	1	3	0	3
Illinois Urbana-Champaign	Official catalogue, 137 course entries	1	0	1	0	2	5	5	7
Georgia Tech	Official catalogue, 273 course entries	1	0	0	0	0	5	2	5
UT Austin	Official catalogue, 178 course entries	0	0	2	0	2	10	6	5
Washington	Official catalogue, 160 course entries	0	0	0	0	2	6	5	5
UPenn	Official catalogue, 143 course entries	8	1	1	0	2	11	4	6
Caltech	Official catalogue, 83 course entries	3	0	1	0	0	5	1	2
US universities naming it	of 12 catalogues read	10 of 12	5 of 12	9 of 12	3 of 12	9 of 12	12 of 12	11 of 12	12 of 12

Not read yet, and why

AUB	Lebanon	No current catalogue found at nine addresses or through the registrar page
Lebanese University	Lebanon	Site shows no readable course pages, even in Chrome
Michigan, Princeton	United States	Sites show a human verification check, even in Chrome
UC Berkeley	United States	Catalogue search returns no results to automated reading; its three course sites are used instead

Cascade to deans

Who receives what, and what they confirm

Faculty or school	Courses	What the dean confirms	Status	Reply by
Faculty of Arts & SciencesDepartment of Computer Science	CSC473, CSC475, CSC612, CSC613, CSC470, CSC476, CSC400, CSC474, CSC463	Check each listed gap against the current syllabus; send the syllabi for the four courses marked for refresh.	Not sent	Provost sets
School of EngineeringEngineering electives	GIN515, GIN517	GIN515 Deep Learning and GIN517 Reinforcement Learning appear in study plans with no catalogue description. Send both syllabi.	Not sent	Provost sets
General EducationOwner to be named by the Provost	GEIT215	Confirm the owner, and whether a unit on AI agents can be added.	Not sent	Provost sets
All other faculties and schoolsMedicine and Medical Sciences, Law and Political Sciences, Architecture and Design, and the rest	·	Their slices have not been run.	Not started	·

Dean sheets

The delta for each course

What each dean receives. USEK's catalogue wording sits beside current course pages from frontier universities, MOOCs and frontier labs. Every quoted topic was checked word for word on the page it links to.

Course	Transformers and LLMs	AI agents	Evaluation	Generative models	Reasoning models	Responsible AI	Production ML
CSC473 Natural Language Processing	Gap	Gap	Gap	·	·	·	·
CSC475 Deep Learning	Gap	Gap	·	Gap	·	Gap	·
CSC612 Creative Computing with Generative AI and LLMs	Gap	Gap	Gap	Covered	·	·	·
CSC613 Emerging Technologies in AI	·	Gap	·	·	Gap	·	·
GEIT215 Generative AI Foundations	Covered	Gap	Gap	·	·	·	·
CSC470 Machine Learning	Gap	·	Covered	·	·	Gap	Gap
CSC476 Applications of AI	·	Gap	·	·	·	·	·
CSC400 Professional Ethics	·	·	·	·	·	Gap	·
CSC474 Artificial Intelligence	·	·	·	·	·	·	·
CSC463 Fundamentals of Data Science	·	·	·	·	·	·	·

Status

Compared with

CSC473 · Faculty of Arts & Sciences

Natural Language Processing

Needs a refreshEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“principles, techniques, and tools used for analyzing and processing natural language data”

The description names no topic.

Peers teach now

Frontier universityStanford CS224N NLP with Deep Learningcurrent page

Self-attention and Transformers
Pretraining (Scaling, Systems, Data)
Post-training (RLHF, SFT, DPO)
Agents, Tool Use, and RAG
Large language model benchmarking and evaluation

MOOCHugging Face LLM Coursecurrent page

Fine-tune Large Language Models
Build Reasoning Models

Not in USEK's description

Transformers and attention
Pretraining and post-training of language models
Agents, tool use and retrieval
Evaluating language models

Already aligned

Hands-on final project on a real NLP problem

Market signalMercor (WEF citation)Indeed Hiring Lab

Ask of the departmentDepartment to confirm whether the syllabus already teaches transformers and language models. If not, rebuild the course around them.

CSC475 · Faculty of Arts & Sciences

Deep Learning (with CSC511 at graduate level)

Needs a refreshEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“build and train deep neural networks for various applications using popular frameworks such as TensorFlow, Keras or PyTorch”

The description names tools, not topics.

Peers teach now

Frontier universityStanford CS230 Deep Learning2026

Adversarial Robustness and Generative Models
Technical Foundations of Responsible AI: Interpretability, Fairness, and LLM Governance
Reinforcement Learning and the Role of RL in Modern AI
Frontiers in AI Applications: Agents, RAG, and Multimodality

Frontier universityMIT 6.S191 Introduction to Deep Learningcurrent page

Deep Generative Modeling
Fine-Tune an LLM
Massively Parallel Training

Not in USEK's description

Generative models
Fine-tuning a language model
Agents, retrieval and multimodal systems
Responsible AI: interpretability and fairness

Already aligned

Building and training deep networks in PyTorch or TensorFlow

Market signalCoursera Job Skills Report 2026 Lightcast

Ask of the departmentAdd a generative-models unit and one responsible-AI lecture. Stanford teaches both inside its core deep learning course; MIT teaches generative models.

CSC612 · Faculty of Arts & Sciences

Creative Computing with Generative AI and LLMs

Needs a refreshEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“introduces students to generative artificial intelligence, focusing on Large Language Models (LLMs)”

“from art to innovation”

Framed around creative use, not engineering.

Peers teach now

Frontier universityStanford CS336 Language Modeling from ScratchSpring 2025

Overview, tokenization
Scaling laws
Alignment and Reasoning RL
Convert raw Common Crawl dumps into usable pretraining data.

MOOCDeepLearning.AI short coursescatalogue on 2026-11-11

Context Engineering
Evaluating and Improving Agentic Applications
Spec-Driven Development with Coding Agents

Frontier labAnthropic Academycatalogue on 2026-11-11

Introduction to Model Context Protocol
Building with the Claude API

Not in USEK's description

How language models are built: tokenization, scaling, alignment
Building with model APIs and the Model Context Protocol
Evaluating agentic applications
Context engineering

Already aligned

Practical use of large language models across domains

Market signalMercor (WEF citation)Coursera Job Skills Report 2026 LinkedIn (via CNBC)

Ask of the departmentKeep the creative track. Add an engineering track built on what the labs and DeepLearning.AI now teach.

CSC613 · Faculty of Arts & Sciences

Emerging Technologies in AI

Needs a refreshEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“recent advancements in machine learning, natural language processing, computer vision, and AI ethics and policy”

Names the fields, not the current advances.

Peers teach now

Frontier universityUC Berkeley Advanced LLM AgentsSpring 2025

Inference-time techniques for reasoning
Agentic workflow, tool use, and functional calling
LLMs for code generation and verification
Coding Agents and AI for Vulnerability Detection
Multimodal Autonomous AI Agents

MOOCHugging Face LLM Coursecurrent page

Build Reasoning Models

Not in USEK's description

Reasoning models
AI agents and tool use
AI for code and security

Already aligned

Ethics and policy as part of emerging technology

Market signalPwC AI Jobs Barometer Indeed Hiring Lab

Ask of the departmentA frontier-topics course should change every term. Tie its outline to the watch so it refreshes on a schedule.

GEIT215 · General Education

Generative AI Foundations

Partly alignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“transformers, GANs, VAEs, and diffusion models”

“ethical and legal implications”

General Education course. It already covers the models and their ethics.

Peers teach now

MOOCDeepLearning.AI short coursescatalogue on 2026-11-11

Context Engineering
Evaluating and Improving Agentic Applications

Frontier labAnthropic Academycatalogue on 2026-11-11

AI Fluency: Framework & Foundations
Introduction to subagents

Frontier labOpenAI Academycatalogue on 2026-11-11

Apply AI at Work

Not in USEK's description

AI agents and how to direct them
Evaluating what an AI system produces

Already aligned

Transformers, GANs, VAEs and diffusion models
Prompting for text, images, video and audio
Ethical and legal implications, bias and privacy

Market signalLinkedIn (via CNBC)Coursera Job Skills Report 2026

Ask of the departmentIt is the only USEK course description that names transformers and diffusion models. Add a unit on working with AI agents; Anthropic and DeepLearning.AI now run courses on it.

CSC470 · Faculty of Arts & Sciences

Machine Learning

Partly alignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“supervised learning, unsupervised learning, and reinforcement learning”

“Clustering, and anomaly detection”

Peers teach now

Frontier universityStanford CS229 Machine Learningcurrent page

Kernels. SVM.
Neural Networks 2 (backprop)
Transformers

Frontier universityCarnegie Mellon 10-601 Introduction to Machine Learning2026

Transformers, AutoDiff, Pre-training, Fine-Tuning
In-context Learning / Reinforcement Learning: MDPs

Frontier labGoogle Machine Learning Crash Coursecurrent page

Intro to Large Language Models
Production ML systems
Responsible AI

Not in USEK's description

A unit on transformers, pretraining and fine-tuning
Production machine learning systems
Responsible AI

Already aligned

Supervised, unsupervised and reinforcement learning
Clustering and anomaly detection
Model building and evaluation

Market signalLightcast

Ask of the departmentThe core matches. Add the transformer unit that CMU and Stanford now include.

CSC476 · Faculty of Arts & Sciences

Applications of AI

Partly alignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“healthcare, finance, and business”

“natural language processing, computer vision, and reinforcement learning”

Peers teach now

Frontier labOpenAI Academycatalogue on 2026-11-11

Apply AI at Work
Lead AI Adoption

Frontier labAnthropic Academycatalogue on 2026-11-11

Claude Code 101
AI Fluency: Framework & Foundations

MOOCDeepLearning.AI short coursescatalogue on 2026-11-11

AI Coding Workflows: From Cloud to Local

Not in USEK's description

Using AI assistants and agents in daily work
Leading AI adoption in an organization

Already aligned

Sector applications in healthcare, finance and business

Market signalLinkedIn (via CNBC)Indeed Hiring Lab

Ask of the departmentAdd a workplace module. Employers ask for AI fluency, and the frontier labs now run their own courses on it.

CSC400 · Faculty of Arts & Sciences

Professional Ethics

Partly alignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“privacy and data protection, intellectual property and copyright, cybersecurity and cybercrime”

Peers teach now

Frontier universityStanford CS230 Deep Learning2026

Technical Foundations of Responsible AI: Interpretability, Fairness, and LLM Governance

Frontier labGoogle Machine Learning Crash Coursecurrent page

Responsible AI

Frontier labAnthropic Academycatalogue on 2026-11-11

AI Fluency: Framework & Foundations

Not in USEK's description

AI governance
Interpretability and fairness of models

Already aligned

Privacy, intellectual property, cybersecurity, professional codes

Market signalMercor (WEF citation)

Ask of the departmentAdd an AI governance unit. Auditing and governance roles are a named growth area.

CSC474 · Faculty of Arts & Sciences

Artificial Intelligence

AlignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“search algorithms, game-playing algorithms, knowledge representation, reasoning, and planning”

“explore ethical issues related to AI”

Peers teach now

Frontier universityUC Berkeley CS188 Introduction to AIFall 2025

A* Search and Heuristics
Game Trees I
MDPs I
Bayes Nets: Representation
HMMs
ML V: Transformers

MOOCHarvard CS50 AI with Pythoncurrent page

Neural Networks

Not in USEK's description

Probabilistic reasoning (Bayes nets, hidden Markov models), if not already taught

Already aligned

Search, game playing, reasoning and planning
Ethics of AI

Ask of the departmentNo change needed now. Check that probabilistic reasoning is in the syllabus.

CSC463 · Faculty of Arts & Sciences

Fundamentals of Data Science

AlignedEvidence: catalogue description vs course pages

USEK catalogue 2025-26 says

“linear and logistic regression, decision trees, clustering, ensemble methods, and text mining”

Peers teach now

Frontier universityUC Berkeley Data 100Fall 2025

Data Cleaning and EDA
Feature Engineering
Cross-Validation, Regularization
Parameter Inference & Bootstrap

MOOCDeepLearning.AI short coursescatalogue on 2026-11-11

Tabular Foundation Models for Data Science Agents

Not in USEK's description

Statistical inference and the bootstrap, if not already taught

Already aligned

Regression, trees, clustering, ensembles
Overfitting and model choice

Market signalIndeed Hiring Lab

Ask of the departmentNo change needed now. Watch for AI agents in data work; DeepLearning.AI already teaches it.

Skills market

What changed since September

Reading of 11 Nov 2026 against the 11 Sep baseline. Every figure checked on its source page. The shift: AI-exposed jobs pay more while entry-level openings in them shrink, and employers want AI fluency joined to judgment, with proof of skill.

Date	Source	Finding	What it means for graduates
2026-09-17	Indeed Hiring Lab	Advertised pay in AI-exposed US jobs rose faster after ChatGPT: premium 5.7% controlling for occupation mix, 4.7% within the same job title, 2.4% holding seniority mix constant.	In the most AI-exposed occupations the entry-level share of postings fell from 29% to 10% between 2021 and 2026, while the senior share rose from 22% to 47%.	Source
2026-09-24	Indeed Hiring Lab economist survey (Q3)	Panel expects more AI downward pressure on median real wages of college-educated workers than a quarter ago; no meaningful pressure for non-college workers. 30% expect substantial or transformative AI productivity impact (29% last quarter).	Expert expectation, not observed wages.	Source
2026-09-28	Coursera Global Skills Report 2026	Critical thinking and judgment are the top human skills, followed by ethics, governance and complex problem-solving. Critical thinking enrollments up 107% year over year. AI micro-credential enrollments up 122%; micro-credentials overall up 33%.	95% of surveyed employers call micro-credentials a key differentiator between candidates (Coursera Micro-Credentials Impact Report).	Source
2026-09-09	Coursera Project Helix	Coursera and Udemy launch a skills platform to identify, build and validate skills with evidence of proficiency.	Employers moving from course content to verified skills.	Source
2026-10-05	Coursera with Anthropic	10 free Anthropic AI courses launched, with skills listed including model evaluation, fine-tuning, context engineering, code review, agent skills and responsible AI.	Frontier labs now publish the AI curriculum themselves, free, refreshed monthly.	Source
2026-10-06	Coursera and Udemy	30+ new programs that blend AI fluency with human judgment rather than teaching them separately.	Pairing AI with human skills is the new course design norm.	Source
2026-11-03	Indeed Hiring Lab (JOLTS, September data)	US job openings little changed at 7.5 million; quits rate 2.1%, above 2% for the first time since June 2025.	Quiet market: few openings, slow churn.	Source
2026-10-02	Indeed Hiring Lab (September jobs report)	US employers added 29,000 jobs in September; unemployment 4.2% (from 4.1%).		Source
2026-08 (latest on page)	LinkedIn Economic Graph workforce report	US national hiring rose 1.9% in August 2026 and is down 6.5% from a year ago.		Source

September baseline, and what could not be updated

Share of US postings that are AI-related (6.3% in Aug 2026): Indeed's September snapshot text gives no new AI share
WEF Future of Jobs and PwC AI Jobs Barometer: Both sites refuse automated reading (403)
Lebanon and Gulf postings: Employer-count watch stopped 18 Oct; skills extraction from postings not built

Date	Who	Signal	Figure as stated	Courses	Delta
2026-06-15	Mercor (WEF citation)	AI trainers, evaluators, auditors and governance specialists are emerging across sectors	25-35% annual growth in human evaluators and trainers	CSC473, CSC612, CSC400	Not measured yet	Source
2026-02-24	LinkedIn (via CNBC)	AI fluency is now asked for alongside traditional skills in postings	not stated	CSC612, GEIT215, CSC476	Not measured yet	Source
2026	PwC AI Jobs Barometer	Jobs requiring AI skills grow far faster than the market	69% versus 9% overall	CSC613	Not measured yet	Source
2025-07-23	Lightcast	AI skills carry a wage premium beyond IT roles	28% salary premium	CSC475, CSC470	Not measured yet	Source
2026-08-24	Indeed Hiring Lab	AI-related postings at a record share of all postings	6.3% of postings, August 2026, versus 3.3% peak in 2022	CSC473, CSC613, CSC463	Not measured yet	Source
2026-04	Coursera Job Skills Report 2026	Generative AI is the top skill learners enroll in	14 GenAI enrollments per minute	CSC475, CSC612, GEIT215	Not measured yet	Source
2026-07-23	Indeed Hiring Lab	Entry-level hiring is weaker than experienced hiring, and the gap has widened	gap wider than the 2022 baseline	CSC476	Not measured yet	Source

Sources and cadence

Who we watch, and how often

Every source is checked daily, because courses, models and job postings change daily. Only the funding, program-page and news checks run daily today; the rest are being wired in.

Group	Entities	What we read	Proposed cadence	Status today
Leading US universities	MIT, Stanford, Carnegie Mellon, Harvard, Columbia, Cornell, Illinois, Georgia Tech, UT Austin, Washington, UPenn, Caltech; plus Berkeley course sites	Official catalogues; course syllabi	Daily	12 catalogues read 11 Nov
Lebanese universities	LAU, USJ, Balamand, LIU, BAU, NDU, Haigazian, AUST; AUB and the Lebanese University not read	Catalogues; new programs	Daily	8 read 11 Nov, 4 with full descriptions
Gulf and regional	MBZUAI, Khalifa, HBKU, KAUST	Program pages; new degrees	Daily	In the daily watch, stopped 18 Oct
MOOCs	Coursera, DeepLearning.AI, Hugging Face, Harvard CS50, fast.ai	New course titles; skill reports	Daily	Fetched 11 Nov
Frontier labs	Anthropic, OpenAI, Google, NVIDIA	Course catalogues; releases	Daily	Fetched 11 Nov
Leading entities	WEF, LinkedIn, Lightcast, PwC, Indeed Hiring Lab, Coursera, Mercor	Skills and jobs reports	Daily	Baseline 11 Sep only
Job postings	Lebanon, Gulf and global employers	Role counts; skills named	Daily	Counts stopped 18 Oct; skills not built
USEK	University Catalogue; department syllabi	Descriptions; syllabi	Daily check for a new catalogue or syllabus	Catalogue read; syllabi requested through the cascade

How it works

From a peer's catalogue to a dean's reply

Seven steps, in order. Scripts fetch, extract and check. A model may assist only after its route is benchmarked and approved.

1
Capture
Scripts fetch each catalogue and course page and keep a dated, fingerprinted copy.
Built; run 11 Nov
2
Extract
Splits catalogues into courses and tests the same topic rules on USEK and every peer.
Built; rules tightened after hand checks
3
Evidence check
Every quoted topic must appear word for word on its page.
77 quotes, 0 failures
4
Compare
Sets USEK against Lebanese and US peers, topic by topic, then course by course.
Computing slice done
5
Cascade
The Provost sends each dean their courses; departments confirm against syllabi.
Not sent
6
Review
Three outside reviewers check the final version before it leaves the Provost.
Not started
7
Repeat
Runs again on the agreed cadence and shows what changed since the last term.
Awaits cadence decision