Blog
Full-stack legal AI
Technical deep dives from the Bind team on designing and operating an AI-first legal stack: model behavior, policy-driven guardrails, multi-agent orchestration, and the systems infrastructure required for reliable autonomous contracting.
Mar 17, 2026 • Sami Kuikka
Prompt Optimization Shouldn't Require Rewriting Your App
How come we have advanved agent tracing products with minimal setup, but prompt optimization itself needing rewriting your product in new frameworks in order to work. My attempt on showing that it does not need to be the case
Read articleMar 16, 2026 • Sami Kuikka
Building Memory Systems that Learn Across Conversations
How to create AI systems that adapt and improve their performance over time by learning from past interactions. Going over the MemGPT and Letta AI frameworks approach on bulding agents with memory systems that evolve through conversations.
Read articleDec 22, 2025 • Sami Kuikka
Automated RAG Evaluation without Human Labels (Ragas in Practice)
Learning and building automated evaluation system with Ragas. Can LLMs automatically evaluate RAG systems? Are human labels always needed?
Read articleDec 9, 2025 • Sami Kuikka
Shrinking Embeddings: Real Results with Mathryoshka Models
Exploring the practical benefits of the Matryoshka Representation Learning technique through my own experiment. Why some models can be shrunk without losing performance, how it works, and does it always work?
Read articleNov 27, 2025 • Sami Kuikka
HyDE: Weaponizing Hallucinations for Better Retrieval
Explaining and building real use case for Hypothetical Document Embeddings (HyDE) to showcase that hallucinations are not always bad
Read articleNov 24, 2025 • Sami Kuikka
Why Early Tokens Matter, and How Tree-of-Thoughts Tries to Fix It
Why LLMs are sensitive to their early tokens, and where prompting techniques like Chain-of-Thought and Tree-of-Thought get their power. This is my attempt to understand why early tokens matter, what ToT does behind the scenes, and whether it’s worth using outside of toy projects.
Read articleNov 17, 2025 • Sami Kuikka
Teaching LLM to improve its own prompts
Building a GEPA-style prompt optimizer from scratch to tune DSPy chains with math-backed intuition.
Read article