--- title: "Tebogo Tseka (@Tebza) | Hashnode" description: "Building with AWSomeness. Read the latest articles by Tebogo Tseka on Hashnode." url: "https://hashnode.com/@Tebza" --- @Tebza # Tebogo Tseka @Tebza · Joined June 2023 Building with AWSomeness ### [How I Run Over 20 AI Agents Locally and Deploy One to Production at a Time](https://tebza.hashnode.dev/how-i-run-over-20-ai-agents-locally-and-deploy-one-to-production-at-a-time?utm_source=hashnode&utm_medium=feed) This article was originally published on tebogosacloud.blog. I have over 20 AI agents. Only one is in production. That is not a constraint. It is a strategy. A system with one excellent production ag 10 min read · [tebza.hashnode.dev](https://tebza.hashnode.dev/how-i-run-over-20-ai-agents-locally-and-deploy-one-to-production-at-a-time?utm_source=hashnode&utm_medium=feed) [Apr 13](https://hashnode.com/posts/how-i-run-over-20-ai-agents-locally-and-deploy-one-to-production-at-a-time/69dd0baa3000edd30fdf321c) ### [The Missing Test Suite: Why AI Projects Fail Before Production](https://tebza.hashnode.dev/the-missing-test-suite-why-ai-projects-fail-before-production?utm_source=hashnode&utm_medium=feed) Most AI projects never reach production. The missing piece is prompt testing with the same rigour as TDD. 18 min read · [tebza.hashnode.dev](https://tebza.hashnode.dev/the-missing-test-suite-why-ai-projects-fail-before-production?utm_source=hashnode&utm_medium=feed) [Apr 2](https://hashnode.com/posts/the-missing-test-suite-why-ai-projects-fail-before-production/69ce811e0ff860b6deecca37) ### [Building an LLM Judge That Doesn't Lie to You](https://tebza.hashnode.dev/building-an-llm-judge-that-doesnt-lie-to-you?utm_source=hashnode&utm_medium=feed) Structural guardrails, multimodal inputs, and a fixed-weight violation catalogue for trustworthy AI evaluation 9 min read · [tebza.hashnode.dev](https://tebza.hashnode.dev/building-an-llm-judge-that-doesnt-lie-to-you?utm_source=hashnode&utm_medium=feed) [Mar 31](https://hashnode.com/posts/building-an-llm-judge-that-doesnt-lie-to-you/69cbd5ea71599e7829dd54e3) ### [5 Models, 467 Actions, 1 Winner — What We Learned Comparing LLMs on Real Code Generation](https://tebza.hashnode.dev/5-models-467-actions-1-winner-what-we-learned-comparing-llms-on-real-code-generation?utm_source=hashnode&utm_medium=feed) Testing Claude, Kimi, DeepSeek on 467 real code generation tasks. The results changed how we think about model selection. 10 min read · [tebza.hashnode.dev](https://tebza.hashnode.dev/5-models-467-actions-1-winner-what-we-learned-comparing-llms-on-real-code-generation?utm_source=hashnode&utm_medium=feed) [2](https://hashnode.com/posts/5-models-467-actions-1-winner-what-we-learned-comparing-llms-on-real-code-generation/69cac9405e4a2652e7fde674) [Mar 30](https://hashnode.com/posts/5-models-467-actions-1-winner-what-we-learned-comparing-llms-on-real-code-generation/69cac9405e4a2652e7fde674) ### [Beyond Text: How We Built an Evaluation Framework for Multi-File AI Outputs](https://tebza.hashnode.dev/beyond-text-how-we-built-an-evaluation-framework-for-multi-file-ai-outputs?utm_source=hashnode&utm_medium=feed) A 4-layer evaluation framework for scoring AI-generated multi-file artifacts using a violation-deduction model 9 min read · [tebza.hashnode.dev](https://tebza.hashnode.dev/beyond-text-how-we-built-an-evaluation-framework-for-multi-file-ai-outputs?utm_source=hashnode&utm_medium=feed) [Mar 30](https://hashnode.com/posts/beyond-text-how-we-built-an-evaluation-framework-for-multi-file-ai-outputs/69cab52aef5f6f725f825fd7)