Software
Resources
Books and research papers that have shaped how I think about software.
Books
Designing Data-Intensive Applications
Martin KleppmannThe definitive guide to distributed systems, data modeling, and the trade-offs behind modern architectures.
Clean Code
Robert C. MartinPrinciples and patterns for writing readable, maintainable software that stands the test of time.
The Pragmatic Programmer
David Thomas & Andrew HuntTimeless advice on craftsmanship, from debugging and estimation to career management.
Structure and Interpretation of Computer Programs
Harold Abelson & Gerald Jay SussmanA foundational text on computation, abstraction, and the nature of programming itself.
A Philosophy of Software Design
John OusterhoutA concise take on managing complexity — deep modules, tactical vs. strategic programming.
Systems Performance
Brendan GreggEnterprise-grade performance analysis methodologies for Linux and cloud environments.
Research Papers
Attention Is All You Need
Vaswani et al., 2017The transformer architecture that started it all. Introduced self-attention as a replacement for recurrence, enabling the parallelism that makes modern LLMs possible.
LoRA: Low-Rank Adaptation of Large Language Models
Hu et al., 2021A parameter-efficient fine-tuning method that freezes pretrained weights and injects trainable low-rank matrices — dramatically reducing the cost of adapting large models.
QLoRA: Efficient Finetuning of Quantized Language Models
Dettmers et al., 2023Combines 4-bit quantization with LoRA to fine-tune 65B parameter models on a single GPU. Makes large model adaptation accessible on consumer hardware.
Language Models are Few-Shot Learners
Brown et al., 2020The GPT-3 paper demonstrating that scaling language models produces strong few-shot learners without task-specific fine-tuning, reshaping how we think about model capabilities.
Training Language Models to Follow Instructions with Human Feedback
Ouyang et al., 2022The InstructGPT paper introducing RLHF — aligning language models with human intent through reinforcement learning from human feedback.
Constitutional AI: Harmlessness from AI Feedback
Bai et al., 2022Anthropic's approach to training helpful and harmless AI systems using AI-generated feedback guided by a set of principles, reducing reliance on human labels.