VIAM LABS
Research

Work from the lab

We study agentic systems, frontier model behaviour, and AI product design — publishing findings openly and building tools along the way.

Research focus

What we study

Four areas where we run experiments, publish findings, and build tools.

Agentic AI

We study how language models can plan, reason, and act across multi-step tasks — including tool use, self-correction, and autonomous decision-making.

Multi-step reasoning and planning
Tool-calling reliability
Self-correction and error recovery
Human-in-the-loop agent design
Multi-agent coordination

Model behaviour

We probe frontier models to understand how they work — and where they fail. Our work spans interpretability, alignment, and emergent capability analysis.

Interpretability and mechanistic analysis
Context window utilisation
Emergent capabilities at scale
Alignment and instruction following
Failure mode cataloguing

AI product design

Closing the gap between research findings and production systems. We study evaluation methodologies, system architecture, and UX patterns for AI-native products.

Evaluation design and evals-as-code
AI-native UX patterns
Latency and reliability tradeoffs
Human feedback loops
Monitoring and observability

Applied ML

Practical research on getting the most out of models in production — from fine-tuning and RAG to inference optimisation and retrieval architecture.

Retrieval-augmented generation (RAG)
Fine-tuning and PEFT methods
Semantic chunking strategies
Inference cost optimisation
Embedding and reranking pipelines

Publications

Papers & reports

All findings published openly. More papers added as research matures.

Evaluating tool-calling reliability in agentic pipelines

A systematic study of failure modes when language models invoke external tools across multi-step tasks — with a benchmark suite and mitigation strategies.

AgentsEvalsTool use

Context window utilisation patterns across frontier models

Empirical analysis of how GPT-4o, Claude 3.5, and Gemini 1.5 use their context windows — revealing systematic gaps between stated and effective capacity.

LLMsBenchmarksContext

VIAM-Bench: an open eval suite for multi-agent coordination

An open-source evaluation framework for measuring coordination, task delegation, and error recovery in heterogeneous multi-agent AI systems.

Open sourceAgentsEvals

Join the lab

We're a small team with a long view

We work independently, publish openly, and stay close to the frontier. If you care deeply about hard AI problems, we'd love to hear from you.