Engineering notes

Technical Guides19 May 2026

A High-Scale Guide to Storing AI Models and Datasets

At Zentara, we are constantly pushing the boundaries of connectivity and infrastructure. As we scale our AI initiatives, one question consistently surfaces among our engineering teams: “How do we version these massive LLM models without breaking our development workflow?”. Managing sourc

Adam

Technical Guides12 May 2026

A Strategic Guide to Layer 4 and Layer 7 Load Balancing

Zentara’s unified communications platform requires processing diverse, high-volume workloads, including REST APIs, video streams, and real-time gaming telemetry. Relying on a monolithic load balancing strategy creates inherent bottlenecks. Therefore, optimizing our infrastructure demands a tie

Adam

topic modeling cyber threat intelligence

Technical Guides21 Apr 2026

How to Use Topic Modelling to Decode Thousands of Cyber Threat Signals

Underground forums, dark web marketplaces, and encrypted chat channels generate millions of posts daily, previous research shows only about 20% of that content is actually relevant to Cyber Threat Intelligence (CTI). Zentara Labs research a multi-engine topic modeling pipeline that applies LDA, BERT

Alisha

Technical Guides14 Apr 2026

How to Implement Advanced HAProxy Sticky Sessions for Scalable Websocket Infrastructure

Why WebSocket Load Balancing Breaks Without Sticky Sessions At Zentara, real-time collaboration tools are the heartbeat of our platform. Therefore, a steady WebSocket connection is the foundation of the user experience. However, scaling stateful protocols like WebSockets behind a high-performance lo

Adam

Technical Guides07 Apr 2026

A Recipe for Langchain LLM Observability using Langfuse

When a standard app fails, you check the stack trace. But when an agent fails, you don’t know what to do. Is the model stuck in a recursive loop? Did it misunderstand the tool output? Is the “Critique Agent” currently hallucinating a list of reasons why the “Research Agent&#8

Adam

Implementing Advanced Sticky Sessions with HAProxy Zentara

Technical Guides06 Mar 2026

How to Implement Advanced HAProxy Sticky Sessions for Scalable Websocket Infrastructure

Why WebSocket Load Balancing Breaks Without Sticky Sessions At Zentara, real-time collaboration tools are the heartbeat of our platform. Therefore, a steady WebSocket connection is the foundation of the user experience. However, scaling stateful protocols like WebSockets behind a high-performance lo

Adam

LLM Observability Langchain Langfuse Zentara

AI in Cybersecurity27 Feb 2026

A Recipe for Langchain LLM Observability Using Langfuse

When a standard app fails, you check the stack trace. But when an agent fails, you don’t know what to do. Is the model stuck in a recursive loop? Did it misunderstand the tool output? Is the “Critique Agent” currently hallucinating a list of reasons why the “Research Agent&#8

Adam

Blog02 Feb 2026

Stop Using “async def” For Your ML Models Unless You Like Latency

A while ago, I was on my way to approve a pull request for a production ready ML internal microservice. The Machine Learning (ML) team had built a working, performant model and wrapped it in FastAPI. The unit tests were all green. During load testing, some of the tests timed out. So I thought maybe

Adam

Technical Guides05 Jan 2026

Guide for Installing Docker on Ubuntu

At Zentara, we understand that true security isn’t just about the policies you write, but the integrity of the infrastructure you run them on, so we’re working hard to eliminate every possible weak link in our software supply chain. To address this, we treated our infrastructure setup like cod

Adam