『How AI Agents Change the Work of an ML Engineer』のカバーアート

How AI Agents Change the Work of an ML Engineer

How AI Agents Change the Work of an ML Engineer

無料で聴く

ポッドキャストの詳細を見る
Dan and Niels Bantilan discuss how AI agents are changing Niels's work on two open-source projects, Flyte and Pandera. Flyte began as an MLOps orchestrator and is evolving into an AI runtime for the code, compute, and execution systems around models and agents. Pandera remains a smaller, community-focused data-validation project. Niels finds agents most useful in mature codebases with strong structure, linters, type checks, and tests. He estimates that his coding velocity has increased at least threefold. Local models handle small fixes, while commercial tools perform better on longer tasks that require broad codebase analysis. Pull requests and code review remain central, with reviewers checking for code smells, security problems, and performance issues. Agents now participate in Niels's debugging loop inside live Kubernetes clusters. Through Flyte's MCP server, an agent can inspect logs, identify an out-of-memory error, update the Flyte configuration, and retry the workload. In one case, an agent found an off-by-one error in tensor loading within five minutes, fixing a model that had been emitting garbage symbols. The experience also exposed a risk: Niels has started skimming the agent's report instead of reconstructing every bug himself. At Union, internal agents have narrow responsibilities and return reviewable artifacts. Nody handles customer requests to change node-pool limits and opens pull requests for engineers to review. Doxy monitors SDK changes and proposes documentation updates. Niels applies the same pattern to PRDs, go-to-market writing, and code examples. Agents should have clear access boundaries and produce work that people can inspect. Niels imagines Flyte letting agents assemble workflows instead of following fixed DAGs. Typed tasks define the available building blocks, while Pydantic Monty safely runs the control-flow code an agent writes. Flyte can move files between pods, route heavy work to suitable compute, and resume a 100-step pipeline at step 98 instead of starting over. Niels sees this as the foundation for an AI runtime that combines agents with training, inference, and reinforcement-learning rollouts. Agents have also made it easier for Niels to maintain Pandera while raising a young family. He is exploring validation schemas for vectors, images, and tensor containers, with Narwhals and LanceDB as possible paths into multimodal data. The design remains open. Pandera's concise plain-text errors work well for agents, while HTML reports may better serve people. Across both projects, Niels sees a continuing human responsibility: understand enough of the system to decide whether an agent's output is worth keeping. Full episode notes Click here to view the episode transcript. Chapters (00:00) - How agents are changing Flyte and Pandera(01:29) - Why agents work best in mature codebases(05:38) - Local models for small fixes, Claude for longer tasks(08:44) - Agents triple coding velocity(11:39) - Flyte MCP keeps Kubernetes out of the debug loop(13:53) - From model training to inference and rollouts(17:17) - Flyte's role in reinforcement-learning workloads(22:08) - Moving tensors between pods and GPUs(23:48) - An off-by-one bug made the model output garbage(25:48) - The risk of losing technical understanding(30:39) - Nody and Doxy: agents with narrow permissions(37:20) - When to move an agent from a terminal into Flyte(45:26) - Agents build execution graphs from typed tools(48:06) - Flyte as a durable AI runtime(51:02) - The case for human ML engineers(52:31) - Extending Pandera to vectors and images(55:10) - Narwhals opens a path to multimodal validation(57:45) - Plain-text errors for agents, HTML reports for people ⠀ Links from the show -------------------- FlytePanderaUnion AIPydantic MontyNarwhalsLanceDBKubernetesRustFS ⠀ Guests ------- Niels Bantilan, Chief Machine Learning Engineer, Union LinkedIn ⠀ Follow the podcast ------------------- LinkedInThreadsInstagramTikTok ⠀ Follow Dan Gerlanc ------------------- XLinkedInThreadsBluesky
adbl_web_anon_alc_button_suppression_t1
まだレビューはありません