Hash Kader is a software engineer and technical writer at Ritza. He contributes hands-on testing of AI tools and developer products to TechStackups, with a focus on what they're actually like to use.
10 articles
Keycloak vs Authentik vs Zitadel, tested with Claude Code: Keycloak ranked first, Zitadel second, Authentik third. What agents know, find, recommend and build.
Herdr vs T3 Code vs Amp, tested on free plans: each ran a software factory turning GitHub issues into a live site. What each tool is and which to use.
SigNoz vs Grafana vs Datadog, tested with Claude Code: SigNoz ranked first, Grafana Cloud second, Datadog third. What agents know, find, recommend and build.
Voxtral TTS vs ElevenLabs, tested: Voxtral 3 against Eleven v4 on emotion, names and numbers, voice cloning and pricing, with every audio clip to hear.
Firecrawl vs Tavily vs Exa vs DeepSeek, tested on 40 searches, 40 page fetches and two site crawls, with pricing, free plans and which to use for each job.
Fish Audio vs ElevenLabs, tested: Drama 3 vs Eleven v4 on voice cloning, emotion, names and numbers, plus what each free plan lets you do.
We tested Orbital vs Hasura and Apollo GraphQL with AI agents. Hasura ranked first; Orbital won recommendations but failed authenticated script access.
An AX audit and review of SpacetimeDB (Spacetime). Can Claude Code find it, recommend it over Convex and Supabase, and build a live app with it?
An agent experience audit of Monospace from Directus, the governed API layer for AI agents. Can Claude Code find it, recommend it, and connect to it over MCP?
Gemini vs ElevenLabs on voice cloning and text to speech: I ran Gemini 3.8 Flash TTS and Eleven v4 through emotion, hard text, long-form, dialogue and cloning tests, with the audio so you can judge.