Professional Certificate in AI Engineering

₦395,000

We accept 3 instalment payments
Paul Adeyemi
AI Engineering Professional
(Lead Instructor)
12 Review
Categories
Intelligence
VOKS Institute Professional Certificate in AI Engineering

Overview

  • Modules 14
  • Projects 5 + Capstone
  • Duration 24 Weeks
  • Skill level Beginner–Applied Intermediate
  • Language English
  • Students 171
  • Assessments Yes
  • Hybrid classes Yes
  • Certification Yes
Course Description

Building useful AI products requires much more than sending a prompt to a language model. Reliable AI applications need well-designed backend services, secure data access, grounded responses, controlled tools, systematic evaluation and dependable deployment.

This programme takes learners through that complete engineering workflow. Starting with Python, APIs and databases, learners progress into model-provider integration, structured outputs, embeddings, vector search, retrieval-augmented generation, AI agents and multimodal applications. They then learn to test AI behaviour, manage security and responsible-use risks, deploy services with containers and cloud infrastructure, and monitor quality, performance and cost.

The focus is practical AI systems engineering: building applications that are testable, observable, maintainable and designed with appropriate safeguards and clearly documented limitations.

Expected Graduate Learning Outcomes

By the end of this programme, learners should be able to:

  • Develop tested Python APIs and integrate SQL databases and AI model providers.

  • Implement structured generation, embeddings, semantic search and grounded retrieval-augmented generation (RAG).

  • Build bounded, permission-aware agent and multimodal workflows.

  • Evaluate AI outputs, retrieval quality, reliability, security, latency and operating cost.

  • Containerize, deploy, monitor and document production-oriented AI applications.

Certification

Certification is based on demonstrated practical work across the programme, including weekly laboratory assignments, five major technical projects and a final capstone. Learners are expected to submit tested, evaluated and documented AI applications and satisfactorily explain and defend their technical decisions, results, assumptions and limitations.

Career Opportunities

Graduates will be prepared for entry-level roles such as:

  • Junior AI Engineer

  • AI Application Developer

  • Python Backend Developer (AI Applications)

  • RAG / LLM Application Developer

  • AI Automation and Agent Workflow Developer

  • Machine Learning Application Support Engineer

  • AI Solutions Developer

Professional Certificate in AI Engineering

A 24-week, 288-hour applied programme focused on building, evaluating and deploying reliable AI applications.

14 Modules24 Weeks288 Learning Hours5 Major Projects + 1 Final Capstone

Programme level: Beginner to applied intermediate. Learning commitment: 144 instructor-led hours and 144 guided/independent hours. The programme focuses on AI systems engineering rather than training frontier models.

  •   MODULE 1 — Introduction to AI Engineering and Modern AI Systems

    Weeks 1 | Recommended classroom allocation: 6 hours per week

    AI engineering versus ML engineering and deep learning; models, API backends, data stores, retrieval, interfaces and monitoring; client-server, synchronous/asynchronous and stateless/stateful architectures; requirements, prototyping, evaluation, security and deployment.

    Required practical output: AI application architecture brief

  •   MODULE 2 — Python Programming for AI Engineering

    Weeks 2–3 | Recommended classroom allocation: 6 hours per week

    Python types, collections, control flow, functions and modules; classes, typing, dataclasses, exceptions and generators; JSON, CSV, files, configuration and secrets; HTTP, requests, retries and asynchronous fundamentals; pytest, logging, formatting, package management and Git.

    Required practical output: Tested Python API integration application

  •   MODULE 3 — Backend Development, APIs and Databases

    Weeks 4–5 | Recommended classroom allocation: 6 hours per week

    HTTP and REST methods, status codes, JSON and headers; FastAPI routing, Pydantic, validation, dependency injection and API documentation; PostgreSQL tables, SQL, relationships, joins, CRUD and SQLAlchemy; routers, services, data-access layers and configuration; authentication, authorization, API tokens, password-hashing concepts; unit and integration tests, mocks.

    Required practical output: PROJECT 1 — Tested FastAPI backend with PostgreSQL integration

  •   MODULE 4 — ML, Neural Networks and Foundation Model Fundamentals

    Weeks 6 | Recommended classroom allocation: 6 hours per week

    Train/validation/test data, overfitting, loss and evaluation; neural-network layers, embeddings, gradients and optimization intuition; transformer tokens, attention, positional information and decoding; LLM pretraining, instruction tuning, context windows, inference and sampling; model selection using privacy, performance, latency, quality and hosted versus self-hosted trade-offs.

    Required practical output: Controlled model comparison and model-selection note

  •   MODULE 5 — LLM APIs, Prompt Engineering and Structured Outputs

    Weeks 7–8 | Recommended classroom allocation: 6 hours per week

    LLM SDKs, authentication, retries, timeouts, rate limits and usage; prompt roles, requirements, few-shot examples, templates and versioning; JSON Schema, Pydantic validation and recovery; grounding, abstention, guardrails and fallbacks; function schemas, tool results and authorization; tokens, caching and usage budgets.

    Required practical output: Structured information extraction service with validated outputs

  •   MODULE 6 — Embeddings, Semantic Search and Vector Databases

    Weeks 9–10 | Recommended classroom allocation: 6 hours per week

    Embedding models, vector representations, cosine similarity and dot products; keyword versus semantic search, ranking and filtering; FAISS, Chroma, pgvector, indexes and HNSW overview; PDF/document parsing, chunking, metadata and source IDs; Recall@k, relevance, hybrid retrieval and reranking.

    Required practical output: PROJECT 2 — Evaluated semantic document search engine

  •   MODULE 7 — Retrieval-Augmented Generation (RAG)

    Weeks 11–12 | Recommended classroom allocation: 6 hours per week

    RAG ingestion, embedding, retrieval and generation; parsing, splitting, indexing and source updates; dense, sparse and hybrid search, reranking and query rewriting; citations, insufficient-evidence responses and hallucination mitigation; retrieval recall, correctness, faithfulness and citation-accuracy evaluation; direct SDK approaches and LangChain/LlamaIndex abstractions.

    Required practical output: PROJECT 3 — Grounded knowledge assistant with retrieval evaluations

  •   MODULE 8 — AI Agents and Tool-Using Systems

    Weeks 13–14 | Recommended classroom allocation: 6 hours per week

    Agents versus workflows, task decomposition and explicit state; tool schemas, execution, permissions and error handling; routes, conditions, human approvals and bounded retries; provider SDKs and LangGraph concepts; multi-agent coordination patterns and limitations; prompt injection, tool boundaries and audit logging; task-success, tool-correctness, cost and latency evaluation.

    Required practical output: Controlled tool-using assistant with tests and permissions

  •   MODULE 9 — Multimodal AI Application Development

    Weeks 15 | Recommended classroom allocation: 6 hours per week

    Text, images, speech and video overview; vision APIs, image descriptions and extraction; transcription, speech synthesis and audio file handling; forms, tables, OCR/layout concepts and validation; uploads, privacy, asynchronous processing and application limits.

    Required practical output: Multimodal document processing prototype

  •   MODULE 10 — AI Evaluation, Testing and Reliability Engineering

    Weeks 16–17 | Recommended classroom allocation: 6 hours per week

    Representative, edge and adversarial evaluation cases; exact-match scoring, rubrics, model-assisted grading and human review; RAG and agent evaluations for retrieval, grounding, tool selection and task success; mocks, unit, integration and end-to-end tests; timeouts, exponential backoff, fallbacks, caching and rate limits; version comparisons and continuous test gates.

    Required practical output: PROJECT 4 — Automated AI quality and reliability evaluation suite

  •   MODULE 11 — AI Application Security and Responsible AI

    Weeks 18 | Recommended classroom allocation: 6 hours per week

    Authentication, authorization, secrets and least privilege; direct and indirect prompt injection and trust boundaries; personally identifiable information, retention, encryption and access controls; unsafe tools, resource exhaustion, output injection and cost abuse; fairness, oversight, transparency and misleading claims; adversarial testing and threat modelling.

    Required practical output: Documented threat model, risk register and mitigations

  •   MODULE 12 — AI Deployment, Cloud Infrastructure and DevOps

    Weeks 19–20 | Recommended classroom allocation: 6 hours per week

    Production services, queues, storage, databases and API providers; Dockerfiles, images, networking and Compose; Azure/AWS/GCP concepts, compute, secrets and managed services; GitHub Actions, tests, releases and rollbacks; FastAPI deployment, HTTPS, hosted applications and vector storage; health checks, scaling, logging and restart strategies.

    Required practical output: Containerized and deployed AI service with automated tests

  •   MODULE 13 — LLMOps, Monitoring and Cost Optimization

    Weeks 21–22 | Recommended classroom allocation: 6 hours per week

    Prompt and model lifecycles, evaluation management and maintenance; traces, request/tool logs, latency, tokens and errors; production regressions and retrieval/agent failure patterns; model routing, context management, caching and batching; streaming, concurrency, timeouts and throughput; prompt/model versioning and rollback; successful-task cost, user experience and reliability KPIs.

    Required practical output: PROJECT 5 — Instrumented and production-oriented AI application

  •   MODULE 14 — AI Engineering Capstone, Portfolio and Career Preparation

    Weeks 23–24 | Recommended classroom allocation: 6 hours per week

    Capstone user needs, architecture, data sources and success criteria; model APIs, backend, retrieval/tools and interfaces; representative evaluation sets, security testing and failure handling; Docker, configurations, monitoring and cost reporting; architecture diagrams, API specifications, README files and limitations; system design, Python, RAG and evaluation interview preparation.

    Required practical output: FINAL CAPSTONE — Complete tested, evaluated and deployed AI application

Practical Projects and Capstone

AssessmentCurriculum areaRequired output
PROJECT 1Backend Development, APIs and DatabasesTested FastAPI backend with PostgreSQL integration
PROJECT 2Embeddings, Semantic Search and Vector DatabasesEvaluated semantic document search engine
PROJECT 3Retrieval-Augmented GenerationGrounded knowledge assistant with retrieval evaluations
PROJECT 4AI Evaluation, Testing and Reliability EngineeringAutomated AI quality and reliability evaluation suite
PROJECT 5LLMOps, Monitoring and Cost OptimizationInstrumented and production-oriented AI application
Final capstoneEnd-to-end independent solutionTechnical report, source code, evaluation, documentation, demonstration and defence

All submissions must state the problem, data provenance, assumptions, technical methods, validation results, limitations and practical interpretation. Projects should use appropriately licensed datasets and avoid publishing credentials or confidential records.

Assessment and Certification

Assessment componentWeight
Weekly laboratory assignments and quizzes15%
Five major technical projects35%
Final capstone project35%
Presentation and technical defence15%
Total100%

Proposed certification requirements: At least 80% attendance; submission of all five projects; at least 60% overall; at least 60% on the capstone; and satisfactory independent technical defence. These are proposed institutional standards subject to approval.

Recommended Teaching and Resources

  • Instructor-led foundations and worked examples, guided implementation with code review, structured laboratory practice and independent projects with feedback.

  • Six instructor-led hours and six guided/independent hours each week; begin capstone preparation before the final week.

  • Use version-controlled source code, reproducible environments, realistic datasets, evaluation checklists and documented rubrics.

Technical Requirements and Primary Technologies

TechnologyApplication in the programme
Python, pytest, GitApplication development, automated tests and source control
FastAPI, PydanticBackend APIs, data validation and service documentation
PostgreSQL, SQLAlchemyRelational data storage and database integration
LLM SDKs and prompt toolsModel-provider integration and structured generation
Embeddings, FAISS, Chroma, pgvectorSemantic search and vector retrieval
RAG frameworks and agent SDKsGrounded knowledge assistants and tool-using workflows
Docker and cloud servicesContainerization, deployment and infrastructure
Monitoring and evaluation toolsQuality, reliability, observability and cost management

Recommended hardware: A modern laptop with 16 GB RAM preferred. Hosted GPU access may be used for selected deep-learning laboratories. Controlled API budgets are recommended for AI Engineering exercises.

Graduate Competencies

  • Build and test Python backend services and connect SQL databases.

  • Integrate language-model APIs and validate structured AI outputs.

  • Implement embeddings, semantic search and evaluated RAG workflows.

  • Develop bounded, permission-aware agent and multimodal application workflows.

  • Evaluate AI quality, reliability, security, latency and cost.

  • Containerize, deploy, monitor and document AI applications.

Industry Benchmark References

These are instructional benchmark resources; this programme is not an official certification from or endorsement by any listed provider. The curriculum should be reviewed at least every six months for API/library changes, industry relevance, resources, graduate project outcomes and employer feedback.

Instructor

Paul Adeyemi    (Lead Instructor)
AI Engineer & AI Solutions Expert

Paul Adeyemi is an AI and software engineering professional who practices what he does best at VOKS Technologies since 2023. Before joining us at VOKS Technologies, he already has 5+ years under his belt.

Reviews

12

3.6 Rating
5 star
3
4 star
6
3 star
3
2 star
0
1 star
0
Chat With Us