Our latest insight
Tips, tutorials, and perspectives from the tekko.id team about the digital and technology world.
Deterministic Simulation Testing in Rust: Mastering Madsim
Eliminate Heisenbugs in distributed systems by using Deterministic Simulation Testing (DST) with Rust and Madsim to create reproducible, bug-free state machines.
Building Memory-Safe Postgres Extensions with Rust and pgrx
Learn how to leverage Rust and the pgrx framework to build high-performance, memory-safe PostgreSQL extensions that outperform PL/pgSQL without the risks of C.
Optimizing Phi-3: A Practical Guide to Fine-Tuning SLMs with Unsloth
Learn how to leverage Unsloth and QLoRA to fine-tune Microsoft's Phi-3 for domain-specific automation, achieving high accuracy with minimal compute.
Edge-Side LLM Inference: Running Local Models with WebLLM and WebGPU
Learn how to deploy high-performance LLMs directly in the browser using WebLLM and MLC-LLM, leveraging WebGPU for privacy-first, zero-latency local AI.
CI/CD for LLMs: Quantifying Regressions with Promptfoo and GitHub Actions
Stop relying on 'vibes' for AI features. Learn how to implement deterministic testing for LLMs using Promptfoo and GitHub Actions to catch regressions in CI.
Verifiable AI: Proving Model Integrity with zkML and EZKL
Discover how to bridge the trust gap in AI using Zero-Knowledge Proofs. This guide explores implementing zkML with EZKL to ensure model integrity in regulated industries.
Carbon-Aware Scheduling: Shifting Kubernetes Workloads for Sustainability
Learn how to integrate the Carbon-Aware SDK with Kubernetes to reduce your cloud's carbon footprint by shifting non-urgent tasks to low-intensity hours.
Continuous Profiling: Using eBPF and Pyroscope for Go Microservices
Learn how to identify CPU hotspots in production Go services using OpenTelemetry, eBPF, and Pyroscope for zero-instrumentation continuous profiling.
Mastering RAG Quality: Evaluation-Driven Development with Phoenix and Giskard
Learn how to move beyond manual spot-checking by implementing Evaluation-Driven Development (EDD) to detect hallucinations and ensure RAG reliability.