High-performance, Rust-accelerated infrastructure for AI agents. Deep dives, benchmarks, and engineering notes from the neul-labs org.https://www.neullabs.com/en-gbhttps://www.neullabs.com/blog/cost-control-for-production-ai-agents/https://www.neullabs.com/blog/cost-control-for-production-ai-agents/Where the money goes in an agent run, how to attribute spend per task rather than per month, and which of caching, batching and budgets to build first.Fri, 11 Sep 2026 00:00:00 GMTagentscostobservabilitycachingrate-limitingringletDipankar Sarkarhttps://www.neullabs.com/blog/database-access-patterns-for-ai-agents/https://www.neullabs.com/blog/database-access-patterns-for-ai-agents/Four ways to give an agent access to a production database, what each one actually defends against, and why the audit trail matters more than the query language.Thu, 10 Sep 2026 00:00:00 GMTagentsdatabasesecurityormaipolicyauditDipankar Sarkarhttps://www.neullabs.com/blog/pyo3-in-production/https://www.neullabs.com/blog/pyo3-in-production/Packaging, ABI choices, GIL handling, error mapping and monkeypatch safety: the decisions behind shipping Rust extensions to Python users.Wed, 09 Sep 2026 00:00:00 GMTrustpyo3pythonpackagingmaturinopen-sourceDipankar Sarkarhttps://www.neullabs.com/blog/the-gil-ceiling-in-python-ai-frameworks/https://www.neullabs.com/blog/the-gil-ceiling-in-python-ai-frameworks/Why LiteLLM, LangGraph and CrewAI stop scaling before the network does, what the GIL actually serialises, and how to tell a GIL wall from an I/O wall.Tue, 08 Sep 2026 00:00:00 GMTpythongilconcurrencyperformancepyo3litellmlanggraphDipankar Sarkarhttps://www.neullabs.com/blog/notes-from-shipping-28-projects/https://www.neullabs.com/blog/notes-from-shipping-28-projects/What we learned shipping 28 MIT-licensed projects: agent runtimes, Rust accelerators, dev tools and governance, and which themes actually turned out to hold.Fri, 05 Jun 2026 00:00:00 GMTmetaopen-sourceneul-labsagentic-infrastructureDipankar Sarkarhttps://www.neullabs.com/blog/drop-in-compatibility-as-a-design-principle/https://www.neullabs.com/blog/drop-in-compatibility-as-a-design-principle/Why most of the Neul Labs portfolio refuses to change your config, your CLI or your imports: the rule, the trade-offs, and the cases where we break it.Fri, 05 Jun 2026 00:00:00 GMTdesigndrop-inrjestrpytestrninjafast-litellmfast-langgraphopen-sourceDipankar Sarkarhttps://www.neullabs.com/blog/agentic-infrastructure-stack/https://www.neullabs.com/blog/agentic-infrastructure-stack/A reference architecture for an agentic stack: workspace runtime, memory, ORM policy, orchestration and payments, built from six Neul Labs primitives.Fri, 05 Jun 2026 00:00:00 GMTagentsarchitectureagentvfsbratormaimcp-paymemorgfastagenticagentic-infrastructureDipankar Sarkarhttps://www.neullabs.com/blog/why-rust-for-ai-infrastructure/https://www.neullabs.com/blog/why-rust-for-ai-infrastructure/The GIL and single-threaded serialization bottleneck production AI agents. Rust with PyO3 gives drop-in replacements for LiteLLM, LangGraph and CrewAI.Sat, 04 Apr 2026 00:00:00 GMTrustpythonperformancepyo3ai-agentsDipankar Sarkarhttps://www.neullabs.com/blog/langgraph-checkpoint-performance/https://www.neullabs.com/blog/langgraph-checkpoint-performance/Checkpoint serialization dominates production LangGraph runs. fast-langgraph uses Rust to cut write time 737x, read time 151x and end-to-end time 2.8x.Sat, 04 Apr 2026 00:00:00 GMTlanggraphbenchmarksrustperformancecheckpointingDipankar Sarkarhttps://www.neullabs.com/blog/multi-agent-orchestration-with-brat/https://www.neullabs.com/blog/multi-agent-orchestration-with-brat/brat is an open-source multi-agent harness that coordinates Claude Code, Aider and Codex on one codebase with crash-safe state, merge queues and a dashboard.Sat, 04 Apr 2026 00:00:00 GMTagentsorchestrationbratclaude-codeaidercodexDipankar Sarkarhttps://www.neullabs.com/blog/brokerless-task-queues-python/https://www.neullabs.com/blog/brokerless-task-queues-python/fastworker is a brokerless Python task queue that removes Redis and RabbitMQ. Priority queues, auto-discovery, distributed workers and a built-in dashboard.Sat, 04 Apr 2026 00:00:00 GMTpythonfastworkercelery-alternativetask-queueredisDipankar Sarkar