Learn the work behind Data, AI & Forward Deployed Engineering
Practical explanations, career decisions and reproducible workflows. Read the reasoning, inspect the evidence and follow the next skill into a real programme.
Define a rollback rule after a model change
Rollback decisions become political when thresholds are invented after a bad graph appears. Write the rule, window, data source and authority before exposing the new model route.
Define active users before calculating DAU and MAU
Define the qualifying behavior, identity, time window and exclusions before counting active users. A page load, heartbeat and completed report represent different levels of product use; treating them as interchangeable c
Define an SLO for successful AI task completion
A healthy API can return fast, syntactically valid failures. For an AI workflow, availability becomes meaningful only when a user receives a correct and authorized outcome within the promised time. That end-to-end event
Define latency budgets for retrieval, generation and tools
An end-to-end latency target tells users what to expect, but it does not tell engineers which stage consumed the budget. Split the objective across retrieval, generation, tools and orchestration before optimization begin
Define operational ownership before client handover
A runbook without owners is reference material, not handover. AI services divide responsibility across platform health, task quality, source data, integration contracts and business decisions. One generic “support team”
Delete a source document from every retrieval index
Removing a file from source storage does not remove its chunks from lexical search, vector collections, caches or evaluation snapshots. A deletion workflow needs lineage and proof that active retrieval can no longer retu
Demonstrate a broken workflow before proposing AI
If the problem is duplicate events or missing ownership, an LLM may add cost without fixing it. Show the failure and baseline first.
Deploy a service with a health endpoint
A health endpoint lets platforms and operators see whether an instance can run and accept work. It should be fast, minimal and free of secrets.
Describe synthetic portfolio data honestly on a resume
Describe a synthetic portfolio project by naming the problem, the work you performed and the result you actually verified. State that the data is synthetic. Do not convert a simulated workflow into a claim of real client
Design a capstone demo that survives difficult questions
A capstone demo should make the project’s decision and evidence easier to inspect. A smooth interface is useful, but reviewers also need to see provenance, baseline, evaluation, failure and operating boundaries.
Design a circuit breaker around unreliable tools
When a dependency is failing, retries can turn a partial outage into a system-wide one. A circuit breaker stops calls after a measured failure condition and gives the dependency time to recover.
Design a feature dictionary with owners and units
A feature name rarely captures enough meaning to reproduce it. spend30d needs currency, grain, time window, refund policy, availability and an owner who can explain changes.
Not sure which programme fits?
Tell us your background and we will map it to the right entry point — including saying so when a cheaper programme is the better fit. A counsellor replies within one working day.