AI Engineering for Backend Developers
Integrate LLM-based features into a Java/Spring backend reliably, measurably and securely.
This program is planned and not scheduled yet. No date, price or cohort has been announced. If you register interest, we contact you when it is scheduled or to discuss an individual format.
Objective
The program teaches a backend engineer to build AI features with the same engineering discipline as any service: LLM API integration, schema-validated tool calling, RAG, evaluation and regression tests, prompt injection defense, privacy, and cost and latency budgets. The focus is reliable backend integration, not model research.
Who it's for
- Middle and senior Java/Spring backend engineers who must add an AI feature to a product
- Tech leads who want to turn an LLM prototype into a dependable service
- Architects who want to understand RAG and tool calling architecture
- Teams that need to measure the quality and security of an AI feature
Who it's not for
- Those who want to learn model training, fine-tuning or ML research
- Those still learning backend fundamentals (HTTP APIs, SQL, Spring Boot)
- Those who just want to use a ready-made AI product; the program targets engineering integration
Prerequisites
- Experience writing REST APIs with Java and Spring Boot
- Working knowledge of SQL and databases
- Basics of JSON and JSON Schema
- A habit of writing tests and working with CI
- Having sent a simple request to an LLM API (desirable)
What you will be able to do
What you will be able to do
- Integrate an LLM API into a Spring service with timeouts, retries and error handling
- Build tool calling with schema validation and never trust arguments returned by the model
- Build a RAG pipeline: chunking, embedding, retrieval and grounding
- Write an evaluation set and regression tests for an AI feature
- Reduce prompt injection and data leakage risk with guardrails
- Monitor cost, latency and quality of AI calls and set budgets
- Limit personal data sent to a model and keep an audit trail
No education program can guarantee employment or a promotion.
Curriculum
Curriculum
- 01
LLM API integration
Call a model API as reliably as any other external dependency.
- Request/response structure and tokens
- Timeouts, retries, rate limits and circuit breakers
- Streaming responses and backpressure
- Abstraction to allow changing providers
- Handling non-deterministic responses
Hands-on lab: An LLM client adapter in a Spring Boot service with retries, timeouts and an error mapping.
- 02
Tool/function calling and schema validation
Let a model invoke backend operations while keeping control.
- Tool description and JSON Schema
- Validating and rejecting model-supplied arguments
- Authorisation: the user's rights, not the model's
- Side-effecting operations and a confirmation step
- Multi-step tool loops and limits
Hands-on lab: A simple agent with order status and refund tools, with schema validation and authorisation checks.
- 03
Retrieval-augmented generation
Ground answers in the organisation's own data and cite sources.
- Chunking strategies and the size trade-off
- Embeddings and vector search
- Hybrid search and re-ranking
- Grounding and source citation
- Knowing when not to answer: not enough context
Hands-on lab: RAG over a document set: indexing with PostgreSQL pgvector, retrieval and an answer that cites sources.
- 04
Evaluation and regression tests
Measure AI feature quality with repeatable checks, not impressions.
- Golden datasets and scenario selection
- Measuring retrieval and generation quality separately
- Rule-based checks and the limits of model-as-judge
- Regression checks in CI
- Effect of prompt and model changes
Hands-on lab: A 30-example evaluation set for the RAG lab and a regression test running in CI.
- 05
Guardrails and prompt injection defense
Stop untrusted text from changing the system's behaviour.
- Direct and indirect prompt injection
- Validating input and output
- Least privilege for tools
- Output sanitisation and rendering risk
- Checking against the OWASP LLM risk list
Hands-on lab: Add a malicious document to the RAG lab, reproduce the attack and test the defenses.
- 06
Privacy and data handling
Stay in control of what is sent to a model and what is stored.
- Masking personal data
- Retention periods and log policy
- Provider terms and data residency questions
- Audit trail and traceability
- User consent and where legal advice belongs
Hands-on lab: A filter that masks personal fields in input and an audit log design.
- 07
Cost and latency budgets
Define the limits that make an AI feature sustainable in production.
- Calculating token cost
- Caching: exact and semantic
- Routing to a smaller model and fallback
- Asynchronous processing and queues
- Per-user limits
Hands-on lab: A cost and latency model: a table of assumptions and a limit mechanism.
- 08
Observability of AI features
See what answered, why, and at what cost.
- Traces: request, retrieval, model call, tool
- Metrics: latency, errors, tokens, refusal rate
- Balancing prompt and response logging with privacy
- Quality signals and user feedback
- Alerting and the SLO concept
Hands-on lab: Observe a RAG service with OpenTelemetry traces and a Grafana dashboard.
Practice assignments
- An LLM client adapter and its tests
- A tool calling service with schema validation
- A RAG pipeline with answers that cite sources
- An evaluation set and a CI regression check
- A prompt injection attack and defense report
Portfolio deliverables
- A sample Spring Boot RAG service project
- An evaluation set and CI check
- A threat model and guardrails document
- A cost and latency budget table
- An observability dashboard configuration
How learning is assessed
- Project review: feedback on code, tests and the evaluation set
- Defense of security and privacy decisions
- Written analysis of cost and quality trade-offs (recommendations, not a certificate)
Tools and technologies
- Java 21
- Spring Boot 3
- Spring AI (as an example)
- PostgreSQL / pgvector
- JSON Schema
- OpenTelemetry
- Grafana / Prometheus
- Testcontainers
- JUnit 5
Level progression
- Before
- Middle Java developer level: Spring Boot, SQL and testing experience.
- After
- Senior backend or system design programs; architecture of AI features.
- Length and format
- Length and format are not set yet; the program is planned and not scheduled.
Frequently asked questions
Frequently asked questions
Is this program available now?
No. It is planned and not scheduled. No dates or prices are announced. You can register interest.
Do I need machine learning knowledge?
No. The program is about backend integration, testing and security, not model training.
Is it tied to a specific model provider?
No. The approach is provider-neutral. Code samples may use one provider, but the adapter layer is designed for switching.
Does it ensure AI quality?
No. No program can guarantee the correctness of AI answers. It teaches how to measure quality and reduce risk.
Articles related to this program
- AI in backend systemsHow to use AI coding agents safely in a real project
- AI in backend systemsCalling an LLM from a backend service safely
- API securityAPI security for backend engineers with the OWASP API Security Top 10
- System designCaching strategies - cache-aside, write-through, TTLs and stampedes
- ObservabilityObservability: logs, metrics and traces
- AI in backend systemsRAG for backend engineers