InDataAIChroniclesbyVedprakash·14h agoTool Design: One Vague Tool Cost Me More Than Ten Sharp OnesPart 2 of 10 in “Ten Disciplines of AI Engineering”. Previous: Part 1, Context Engineering. Next: Part 3, Memory Architecture.
InDataAIChroniclesbyVedprakash·1d agoContext Engineering: The Agent Stopped Guessing the Join When I Gave It LessPart 1 of 10 in “Ten Disciplines of AI Engineering”. Previous: Part 0, the series map. Next: Part 2, Tool Design.
InDataAIChroniclesbyVedprakash·1d agoThe AI Engineering Interview Has Changed. Here Is the Map I Wish I Had.Part 0 of 10 in the series “Ten Disciplines of AI Engineering”. This is the index. Each part below becomes a post with a diagram, a worked…A response icon2A response icon2
InDataAIChroniclesbyVedprakash·3d agoMy Agent Harness Gave a Perfect Score to an Agent That Skipped 300 UpdatesAgents That Load Data, Part 1: one loading task held constant across the Claude Agent SDK, the Snowflake Cortex Agent SDK and Google ADK…
InSnowflake Builders Blog: Data Engineers, App Developers, AI, & Data SciencebyVedprakash·Oct 29The Agent Is Now a Pipeline Step: Keeping CoCo Automations and Cloud Agents ReproducibleRunning Snowflake in the Agent Era, Part 4 of 4. An agent on a schedule is a pipeline step, and it needs the same review, versioning, and…
InSnowflake Builders Blog: Data Engineers, App Developers, AI, & Data SciencebyVedprakash·Oct 29Budgets, Quotas, and Monitors: The Cost Controls I Map Before Agents Go WiderRunning Snowflake in the Agent Era, Part 3 of 4. The cost_center tag from Part 1 finally earns its keep, and I learn which control actually…A response icon1A response icon1
InSnowflake Builders Blog: Data Engineers, App Developers, AI, & Data SciencebyVedprakash·Oct 26From ORGADMIN to Control Plane: Organization Hub, Organization Users, and Premium Views in PracticeRunning Snowflake in the Agent Era, Part 2 of 4. Part 1 made compute flexible inside one account. This part is about the one account that…
InDataAIChroniclesbyVedprakash·Oct 24Graphs Don’t Kill Prompts. They Give Them Contracts.Agent workflows built as graphs are everywhere this month. The prompt didn’t die. It moved into nodes with inputs, outputs and a verifier…
InDataAIChroniclesbyVedprakash·Oct 22Decision Models: When a Probability Beats a ParagraphA new model category answers typed questions with probabilities instead of text. Here’s how I’d evaluate one before letting it route real…
InSnowflake Builders Blog: Data Engineers, App Developers, AI, & Data SciencebyVedprakash·Oct 22Stop Picking T-Shirt Sizes: What Running Adaptive Compute Taught MeRunning Snowflake in the Agent Era, Part 1 of 4. Two dials replaced four knobs, and the hardest decision moved somewhere I did not expect.A response icon1A response icon1
