Professional programmes
Generative AI
الذكاء الاصطناعي التوليدي
Build reliable generative-AI applications, grounded workflows and responsible intelligent products.
- Theory
- 36 hours
- Practical and lab
- 48 hours
- Assessment and capstone
- 12 hours
- Length
- 3 months
- Planned hours
- 96 hours
- Weekly schedule
- Two 4-hour sessions a week
- Level
- Advanced applied
- Delivery
- On campus, or blended where the approved programme permits.
- Location
- Riyadh
- Next start
- To be confirmed
Overview
Participants finish with a tested portfolio application, evaluation evidence, risk controls, documentation and a professional demonstration—not merely a collection of prompts.
This three-month course develops the ability to design, build, evaluate and govern professional generative-AI applications. Learners progress from model and prompt foundations to API applications, multimodal workflows, embeddings, RAG, tool use, supervised agent patterns, evaluation, safety and deployment, culminating in an end-to-end capstone.
Who it is for
Who should join, and what to bring.
Who should join
- Graduates and final-year students in computing, engineering, business, media or related fields.
- Developers, analysts, product teams, researchers and automation professionals.
- Marketing, content, education, HR, operations and knowledge-management professionals.
- Managers and entrepreneurs evaluating responsible AI products and workflows.
Recommended before you start
- Comfort with computers and professional digital tools.
- Basic Python recommended for API/application labs; an approved bridge may be assigned.
- Logical problem-solving, basic data awareness and English technical reading.
- Suitable laptop and approved model/API access.
What you will learn
By the end, you will be able to:
- Frame a generative-AI use case with value, success measures and constraints.
- Design robust prompts and validate structured outputs.
- Build Python/API-based text and multimodal applications.
- Create embedding-based semantic search and evidence-grounded RAG.
- Integrate approved tools/functions with validation and permission boundaries.
- Design supervised agentic workflows with stop conditions and human approval.
- Build evaluation datasets, rubrics and regression tests.
- Identify and mitigate hallucination, bias, privacy, prompt-injection and IP risks.
- Deploy a prototype with secrets, logging, feedback and version control.
- Deliver and defend an end-to-end capstone with documentation.
- Use case
- Prompt/application
- Knowledge retrieval
- Tools/workflow
- Evaluation
- Risk controls
- Deployment
- Monitoring
Curriculum
Twelve weeks, three months.
Foundations, Prompting, APIs and Multimodal Workflows
Weeks 1 to 4
- AI/ML/deep-learning and generative-model terminology
- LLM, diffusion and multimodal concepts at a practical level
- Use-case value, constraints, stakeholders and success measures
Lab: use-case canvas and baseline model interaction
- Instruction, context, examples, constraints and output schemas
- Zero/few-shot, role, decomposition and iterative refinement
- Structured JSON/table outputs and validation
Lab: reusable prompt patterns and test cases
- API concepts, authentication, requests, responses and errors
- Python application structure, configuration and secret handling
- Conversation state, token/cost/latency awareness and retries
Lab: controlled text-generation application
- Text summarisation, extraction, classification and transformation
- Image/audio/document inputs and output-quality criteria
- Human review, attribution, IP and brand controls
Mini-project 1: multimodal professional workflow
Tools and projects
What you will work with, and what you will make.
Tools
- Python, notebooks/IDE, Git and environment/secret management.
- Approved commercial and/or open-source model APIs and interfaces.
- Hugging Face Transformers or approved model-learning environment.
- Embedding/vector-search environment and document-processing tools.
- Simple web application framework such as Streamlit or approved alternative.
- Evaluation datasets, rubrics, logging and monitoring tools.
- Approved multimodal generation tools subject to licences and policy.
The final set of tools is confirmed for each batch.
Example projects
- Evidence-grounded institutional knowledge assistant
- Document extraction, summarisation and compliance-review workflow
- Multimodal marketing/content production system with human approval
- Customer-support drafting and quality-assurance assistant
- Education or training content assistant with source validation
- Operations copilot integrating approved tools and structured outputs
The capstone: what you hand in
- Use-case brief, users, success metrics and risk classification.
- Architecture, prompts/configuration, source/retrieval design and tool boundaries.
- Working application with repository and setup instructions.
- Golden test set, evaluation results, failure taxonomy and mitigations.
- Privacy, security, IP, responsible-use and human-oversight statement.
- User guide, technical handover and executive demonstration.
Assessment
Assessed on the work you produce.
- Labs25%Working outputs and validation
- Applied assignments20%Independent workflow design
- Mini-projects20%Integrated AI applications
- Capstone25%Evaluated responsible product
- Presentation/practice10%Defence, disclosure and teamwork
How it is taught
- Concept explanation and live model/application demonstration.
- Guided labs, test-driven iteration and failure analysis.
- Case studies, peer critique and mentor checkpoints.
- Human verification, responsible-AI review and documented evidence.
- Two mini-projects and an integrated capstone.
Certificate wording and issuing details will follow the final TVTC approval. No accreditation is claimed before that approval.
Dates and fees
Start dates and fees, on request.
Next start
To be announced
Enquire now and we will tell you the next start date and schedule.
Fees
Request fee details
We send the fee, and any options for paying, with your enquiry reply.
Questions
Questions people ask.
Prompting is included, but the course goes further into applications, RAG, tools, evaluation, safety and deployment.
Basic Python is recommended for technical labs; diagnostics and approved bridge support may be provided.
Approved commercial and/or open-source platforms may be used. The final stack depends on licences, availability and policy.
Yes—two mini-projects and a professional capstone are recommended.
No. It develops demonstrable capability and portfolio evidence; employment is not guaranteed.
Beyond the course
From learning to a technology career.
Strong performance in labs, projects and capstones can lead to further opportunities with TechnoBerg. They are merit-based, subject to availability, legal eligibility and formal selection, and never guaranteed. See the talent pathway