Customer-obsessed science
Research areas
-
November 16, 202611 min readNew open-source tool turns existing test suites into sets of formal-verification tasks, which catch buggy solutions that tests alone would miss.
-
November 11, 20265 min read
-
-
-
Featured news
-
2026An agent inference can execute immediately through a real-time streaming API or be deferred to a batch API that cloud providers price at approximately half the real-time rate. Existing agent interfaces generally bind a request to one mode before execution. The runtime therefore cannot revise the mode as conditions change. We identify three requirements for per-inference mode scheduling: (i) the agent definition
-
2026Enterprises increasingly want LLM agents to automate standard operating procedures over their software stacks—mapping each step of a procedure to the right call among thousands of internal APIs, without hand-authored step-to-API mappings. The core difficulty is API selection at scale: choosing and composing the correct tools over a long horizon, judged only by whether the overall task succeeds. Current
-
ACM SIGSPATIAL 20262026We present a unified scene understanding system for automated road closure mapping at scale. Our approach detects, segments, and geo-localizes road objects and surfaces from fleet camera image sequences, and conflates them with map segments—all without human intervention. For detecting road closure objects, we adopt a two-stage strategy using single-stage detection and segmentation models: a lightweight
-
DAI 20262026Synthetic benchmarks are increasingly used to evaluate AI agent memory systems, yet while recent work has examined bias in synthetic test collections for information retrieval and RAG evaluation, we find none examining whether document generation strategy affects intrinsic evaluation of agent memory systems, where generated artifacts must encode temporal fact evolution across multi-document sequences. We
-
2026Production LLM agents often waste latency and reliability by regenerating code for the same procedural steps on every request. We replace this inference-time coding loop with an agentic tool-making pipeline that compiles repeated SOP steps into validated, versioned tools before deployment. The tool-maker grounds synthesis in the live environment as it collects execution traces, observes backend schemas
Collaborations
View allWhether you're a faculty member or student, there are number of ways you can engage with Amazon.
View all