Plate A-04 · Programme Catalogue
Three Paths Through AI Practice
Each programme covers a distinct domain of AI work. They are designed to be taken in sequence or individually, depending on what your practice needs next.
Back to HomePlate B-04 · Methodology
How Latentry Programmes Are Structured
Plate-by-plate sequencing
Sessions are organised as plates in a curriculum atlas. Each plate states its dependencies and its destinations. You can see the full map before you begin and track your position throughout.
Practice with real constraints
Exercises use datasets and scenarios that reflect actual conditions in practice — awkward schemas, class imbalance, infrastructure cost limits. The goal is habit formation, not demonstration of ideal conditions.
Post-cohort curriculum review
After each cohort closes, the teaching team reviews the curriculum against current practice. Any session that no longer reflects what practitioners actually encounter is updated before the next intake opens.
Plate C-04-A · Programme 1
Introductory Cohort: Data Handling and Baselines
A nine-week cohort covering the groundwork that decides whether later modelling is worth anything: sourcing data responsibly, understanding schemas, cleaning without quietly destroying signal, splitting data honestly, building sensible baselines and knowing when a simple model is the right answer. Learners work through three datasets of increasing awkwardness and produce short written analyses alongside their code. Suitable for people with basic Python who want disciplined habits early.
What is included
- 18 live sessions with written feedback on each exercise
- Dataset library access (three progressively complex datasets)
- Peer review pairing for three exercises
- Completion record on finishing all assessed work
How it runs
- 1Weeks 1–3: responsible data sourcing, schema analysis, documentation habits
- 2Weeks 4–6: cleaning strategies, signal preservation, honest train/test splitting
- 3Weeks 7–9: baseline selection, evaluation framing, written analysis review
Plate C-04-B · Programme 2
Computer Vision Specialisation
A fifteen-week specialisation covering image data end to end: preprocessing and augmentation, convolutional architectures, object detection and segmentation approaches, transfer learning, evaluation metrics suited to imbalanced visual tasks, and the practical constraints of inference on modest hardware. Learners build three systems including one deployed to a small device or web endpoint. Aimed at those with prior modelling experience.
What is included
- 30 live sessions with GPU credits provided
- Three reviewed builds including one deployed system
- Weekly critique session with cohort peers
- Mentor office hours and written portfolio review
How it runs
- 1Weeks 1–5: image preprocessing, augmentation, convolution fundamentals and first build
- 2Weeks 6–10: detection, segmentation, transfer learning and second build
- 3Weeks 11–15: inference constraints, deployment, evaluation and portfolio review
Plate C-04-C · Programme 3
MLOps and Deployment Programme
A nine-month part-time programme focused on everything that happens after a model works on a laptop. Modules cover reproducible environments, data and model versioning, pipeline orchestration, containerisation, serving patterns and latency budgets, monitoring for drift and failure, incident practice, cost management and the documentation a maintainer actually needs. The programme runs on a continuous team project with rotating roles, reviewed at five checkpoints by working engineers. Suitable for developers with production experience.
What is included
- Weekly live sessions and cloud credits
- Five project checkpoints reviewed by working engineers
- On-call simulation exercises and mentoring
- Architecture writing workshop and detailed final assessment
How it runs
- 1Months 1–3: reproducible environments, versioning, pipeline orchestration
- 2Months 4–6: containerisation, serving patterns, latency budgets, checkpoint 3
- 3Months 7–9: drift monitoring, incident practice, cost management, final assessment
Plate D-04 · Decision Guide
Which Programme Is Right for You?
Use this overview to match your current experience level and learning goal to the right path.
| Feature | Data Cohort RM 480 |
CV Spec. RM 1,960 ★ Popular |
MLOps Prog. RM 4,380 |
|---|---|---|---|
| Entry level | Basic Python | Prior modelling exp. | Production exp. |
| Duration | 9 weeks | 15 weeks | 9 months |
| Weekly commitment | 7 hrs | 10–12 hrs | 12–15 hrs |
| Live sessions | 18 | 30 | Weekly |
| GPU credits | — | — | |
| Cloud credits | — | — | |
| Engineering checkpoint reviews | — | — | 5 reviews |
| Portfolio / written output | Completion record | Portfolio review | Architecture doc + assessment |
| Best for | Getting data right first | Building visual AI systems | Keeping models in production |
Plate E-04 · Standards
Professional Standards Across All Programmes
Data Privacy
Learner data is held only for programme administration. Datasets used in exercises are either synthetic or shared under open licences with clear attribution.
Feedback Turnaround
Exercise submissions receive written feedback within five business days. MLOps checkpoint reviews are returned within seven business days with a written report.
Learner Support
Administrative queries sent to [email protected] are handled within two business days. Technical questions are addressed in live sessions and office hours.
Version-Controlled Content
Curriculum materials and exercise datasets are version-controlled. Changes between cohorts are documented. Enrolled participants continue with the version they started.
Assessment Integrity
Completion records reflect actual assessed work. MLOps Programme checkpoints are reviewed by engineers who are not affiliated with the teaching team for that cohort.
Secure Learning Platform
Session recordings and submissions are stored on a platform with access controls. Access is tied to enrolment status and disabled on withdrawal or programme completion.
Plate F-04 · Pricing
Programme Fees
Data Cohort
RM 480
per enrolment · 9 weeks
- 18 live sessions
- Written feedback
- Dataset library
- Peer review pairing
- Completion record
CV Specialisation · Popular
RM 1,960
per enrolment · 15 weeks
- 30 live sessions
- GPU credits
- Three reviewed builds
- Weekly critique session
- Mentor office hours
- Portfolio review
MLOps Programme
RM 4,380
per enrolment · 9 months
- Weekly sessions
- Cloud credits
- 5 engineering reviews
- On-call simulation
- Architecture writing workshop
- Detailed final assessment
Plate G-04 · Enrolment
Not Sure Where to Start?
Tell us where you are in your practice and what you want to be able to do next. We will suggest the most appropriate path without any pressure to commit immediately.
Get in Touch