Dr. Ling Liu

Professor; Director, Distributed Data Intensive Systems Lab (DiSL)

School of Computer Science, Georgia Institute of Technology

Ling Liu directs DiSL, examining performance, availability, security, privacy, and trust in large-scale distributed and cloud systems. She served as Editor-in-Chief of IEEE Transactions on Services Computing (2013–2016) and has published over 300 papers on distributed data-intensive systems — the substrate a multi-tenant agent runtime has to get right at scale.

IEEE Fellow

PhD, Tilburg University

IEEE Computer Society Technical Achievement Award, 2012

Former EiC, IEEE Transactions on Services Computing

the partnership

The runtime underneath everyone else’s layer.

Semantics, authority, and collective behavior all have to actually execute somewhere — at scale, across edge, regional, and cloud tiers, for thousands of concurrent agents that don’t behave like a predictable cloud workload. That’s the substrate Ling Liu’s group has spent decades getting right for distributed and cloud systems generally, now applied to agent runtimes.

The work makes the runtime itself a first-class partner in trustworthy agent infrastructure: the place where policy is enforced, tenants are isolated, and fleet behavior stays legible as conditions change.

the research

Operational trust, where the work happens.

Three runtime directions translate Ling Liu’s distributed-systems research into the operational backbone of Algedonic’s Secure Execution Fabric.

DIRECTION 01

Scalable Distributed Policy Enforcement

Where should policy enforcement live across edge, regional, and cloud tiers so it stays consistent, auditable, and available without sacrificing latency?

Secure Execution Fabric · sidecars · regional runtime cells

DIRECTION 02

Elastic Orchestration of Multi-Tenant Agent Runtimes

How can shared agent infrastructure isolate tenants, absorb unpredictable spikes, and preserve latency guarantees when one fleet behaves badly?

Secure Execution Fabric · tenant isolation · elastic scheduling

DIRECTION 03

Detecting Safety Drift in Long-Running Agent Fleets

Monitor the live fleet against its certified baseline and flag aggregate safety drift before it becomes an incident.

fleet observability · certified baselines · continuous safety

also mapped to this lab

Evidence that travels with the runtime.

Privacy-Preserving Runtime Governance

Can an enterprise prove policy compliance without exposing prompts, memory, sensitive data, or proprietary reasoning? Building confidential telemetry and Data Trust Zones with ZK and TEE-based approaches at distributed-systems scale.

Evidence architecture · Data Trust Zones · confidential telemetry · ZK/TEE

Verifiable Distributed Agent Transactions

How can agent actions and delegations be cryptographically proven across organizational boundaries? Apply distributed consensus and ledger techniques so every transaction carries a provable evidence chain.

Dual ledger · evidence chain · provenance · signed A2A

why it matters

"A governance model is only as trustworthy as the runtime that actually has to execute it. Policy placement, multi-tenant isolation, and drift detection at scale are distributed-systems problems, not security problems wearing a different hat — so instead of bolting our enforcement logic onto infrastructure we hoped would hold up, we went to the person who's spent her career making distributed systems hold up under exactly this kind of pressure."

— Sandeep Gopisetty, Founder & CEO, Algedonic

← Back to research portfolio