Tenzaa's Tech Blog
This site is dedicated to sharing IT-related technologies.
Building a Resilient Saga Pattern with AWS SQS and Local SQLite State Machines Building a Resilient Saga Pattern with AWS SQS and Local SQLite State Machines
A critical business process involving multi-step data processing across different domains was causing recurring data con
2023-10-27 Architecture
Monorepo SQL AWS SQS Saga Pattern BDD SQLite
Automating Time Series Vector Embedding Pipelines with Tekton and Weaviate under Phoenix Observability Automating Time Series Vector Embedding Pipelines with Tekton and Weaviate under Phoenix Observability
The operational bottleneck wasn’t model training; it was the deployment and validation of time-series embeddings for our
2023-10-27 MLOps
Observability Time Series CI/CD Phoenix Weaviate Tekton Vector Database
Constructing a Polyglot Data Ingestion Pipeline with Elixir and Rust for Apache Hudi on AWS Constructing a Polyglot Data Ingestion Pipeline with Elixir and Rust for Apache Hudi on AWS
The monolith was failing. Our IoT analytics platform, responsible for ingesting real-time sensor data from millions of d
2023-10-27 Data Engineering
Apache Hudi gRPC Turbopack Elixir AWS Rust Tonic
Implementing a Dynamic Request Batching Layer in Go for a Docker Swarm-Hosted Transformer Model Implementing a Dynamic Request Batching Layer in Go for a Docker Swarm-Hosted Transformer Model
The core problem wasn’t a complex algorithm, but a simple, brutal mismatch of throughput. We had a sentence-transformer
2023-10-27 Architecture
Performance Microservices Go-Gin Hugging Face Transformers Docker Swarm
Implementing a High-Throughput Polyglot Data Pipeline with Phoenix, Rust, and AWS SQS Implementing a High-Throughput Polyglot Data Pipeline with Phoenix, Rust, and AWS SQS
The initial system bottleneck was a Phoenix controller action that timed out. Users uploaded multi-gigabyte CSV files fo
2023-10-27 System Design
AWS SQS Sentry Pandas Phoenix Rust
Constructing a CDC-Based Vector Synchronization Pipeline from MariaDB to Weaviate Constructing a CDC-Based Vector Synchronization Pipeline from MariaDB to Weaviate
Our primary product catalog, a critical asset, resides in a well-structured MariaDB instance. While excellent for transa
2023-10-27 Data Engineering
CDC Debezium Kafka Memcached Weaviate Jupyter MariaDB
6 / 34