AI platform and agent infrastructure
Building platform foundations that help teams move from isolated AI experiments to reusable, observable, and production-ready capabilities.
AI Platform · Enterprise Data · Engineering Leadership
Jiankun Liu
Engineering leader building AI platforms, enterprise data systems, and practical AI products.
I help teams turn fragmented data, tools, and workflows into reliable AI-enabled products. My work sits at the intersection of AI platforms, enterprise data, backend systems, search and retrieval, and engineering execution.
AI Platform
Practical AI systems
Enterprise Data
Governed data access
Platforms
Backend foundations
Leadership
Strategy to execution
Profile
Building the infrastructure layer behind practical AI products.
I am an engineering leader with experience building AI-enabled products, enterprise data platforms, backend systems, search and retrieval capabilities, and applied machine learning workflows. I focus on the infrastructure that makes AI useful in real organizations: data access, retrieval, orchestration, governance, observability, and scalable engineering delivery.
Across my work, I care less about AI demos that look impressive once, and more about systems that teams can operate, integrate, and trust in production.
Experience signals
Systems, platforms, and leadership patterns behind practical AI.
Enterprise data access
Designing backend and data access layers that turn fragmented business data into governed, reliable, product-facing systems.
Search, retrieval, and capability resolution
Applying search and retrieval patterns to help users, applications, and agents find the right knowledge, data, tools, and capabilities under real constraints.
Engineering leadership from ambiguity to production
Leading teams through unclear requirements, architectural tradeoffs, delivery pressure, and the operating discipline required to ship durable systems.
Venture
Lattivia
Lattivia is an agent capability search and resolution layer for AI workflows. It helps supervisor agents and AI platforms discover, rank, and select the right capability before execution.
The product is built around a simple thesis: as organizations adopt more APIs, MCP servers, tools, workflows, and sub-agents, the hard problem becomes capability resolution — knowing what the agent should use, under what constraints, with what risk and authorization requirements.
Visit LattiviaFocus
How I create leverage
AI platform strategy
From scattered AI initiatives to shared platform capabilities, operating models, and delivery paths that teams can repeatedly use.
Enterprise data platforms
Turning complex enterprise data, APIs, and business logic into governed services that products and AI workflows can depend on.
Engineering leadership
Creating clarity across product, architecture, execution, and delivery when the problem is ambiguous and the organization is moving fast.
Agentic AI infrastructure
Designing the resolution, context, policy, and observability layers agents need before they can safely and reliably act.
Search and retrieval systems
Using retrieval to connect fragmented knowledge, operational data, APIs, and capabilities into usable product experiences.
AI product execution
Moving AI ideas from prototype to production through architecture, evaluation, deployment, monitoring, and iteration loops.
Perspective
Field notes from building data, AI, and platform systems.
I write about what I have learned building enterprise data platforms, AI-enabled products, search and retrieval systems, and engineering teams. The focus is practical: how ambiguous technology trends become systems that teams can operate, scale, and trust.
Visit PerspectiveContact