Reading list
Outside articles and docs to read alongside the missions. They are written by other people and organisations, open in a new tab, and are not part of the Drillhall curriculum or assessment.
Forward deployed engineering
What the role is and how it differs from product engineering.
- Dev versus Delta: Demystifying engineering roles at Palantir · Palantir blog
The role split between building one platform for many customers and deploying it deeply for one.
- A Day in the Life of a Palantir Forward Deployed Software Engineer · Palantir blog
A first-person look at the day-to-day work with customers.
- Who Wants to Be a Delta? · Palantir blog
What the role asks of the person doing it.
Building with AI
Patterns for LLM-based systems and agents.
- Building effective agents · Anthropic
When to use workflows versus agents, and the simple composable patterns that hold up.
- Prompt engineering overview · Anthropic docs
Where prompting fits and what to try before reaching for fine-tuning.
- Patterns for Building LLM-based Systems & Products · Eugene Yan
Evals, retrieval, guardrails and feedback loops as a pattern catalogue.
- Building A Generative AI Platform · Chip Huyen
The components around the model: context, guardrails, routing, caching, orchestration.
- LLM Powered Autonomous Agents · Lilian Weng
Planning, memory and tool use as the pieces of an agent.
Evaluation
How to know whether an AI feature works.
- Your AI Product Needs Evals · Hamel Husain
A practical case for evals as the loop that makes an AI product improve.
AI risk and security
Threats and risk frameworks for AI systems.
- OWASP Top 10 for Large Language Model Applications · OWASP
The common vulnerability classes in LLM applications, such as prompt injection.
- Prompt injection: What's the worst that can happen? · Simon Willison
Why injection matters once a model can call tools or read untrusted text.
- AI Risk Management Framework · NIST
The framework regulated customers tend to reference when they ask about AI risk.
Running it in production
Deployment, incidents and postmortems.
- Managing Incidents · Google SRE book
Roles, communication and handoffs during an incident.
- Postmortem Culture: Learning from Failure · Google SRE book
Blameless postmortems and what makes them useful.
- The Twelve-Factor App · Heroku
Baseline conventions for config, logs and processes in deployable services.
- Podman documentation · Podman
Reference for the container runtime the labs run on.