Aishat Oluwaseun Idowu·1d agoAI & Intelligent Systems Evaluation Portfolio.Evaluating AI-Generated Responses: Accuracy, Relevance, Completeness and Reliability
Sawera Yaseen·2d agoOne Concept, Many Fields: My Rabbit Hole Into AffordancesMy flatmate and I have a small routine. Most nights we end up talking about our work. Sometimes it’s our research, sometimes something one…A response icon1A response icon1
Ehsan Barkhordar·2d agoAgents wrote their own memory notes, and 93% widened what the user had actually saidA new study tests agent memory outside the scope it was formed in, which is the test most memory evaluations never run.
Priti Solanki·4d agoHow I’d Evaluate AI-Generated College Recommendation LettersWhat one factual error taught me about designing GenAI systems that people can review and trust.
Anastasiia Bakhtoiarova·5d agoReliability isn’t free: what each check costs in a multi-agent pipelineWhere the cost of trusting a multi-agent pipeline lands, and the check I still can’t audit.
Talieh Rohani·6d agoWhat Is an AI Eval? How to Tell If Your AI Feature Actually WorksA plain guide to how AI teams test their products, and why the PM should help write the test. Part of my ongoing series on AI concepts PMs…A response icon1A response icon1
InAI at Lloyds Banking GroupbyAI at Lloyds Banking Group·6d agoVERGE: Building Better Multi-Hop RAG Evaluation for Enterprise KnowledgeIs your RAG really as good as you think it is?
Yedukondalurd·Nov 3Beyond the Chatbot: Building a RAG System That Measures, Improves, and Knows When to Ask a HumanI built an AI investment-research assistant with grounded retrieval, LLM-as-judge evaluation, self-optimizing prompts (DSPy + GEPA), and…
AI Spec Kit·Nov 3Golden Datasets for PMs: How to Build the Test Set Your AI Feature Is Judged ByOriginally published at AI Spec Kit on Substack.
ARMARRAY·Nov 3What Counts as Verifiable Completion in a Mobile App?A $42 expense claim shows why plausible app feedback and verifiable completion are different questions.
