callosum · London ·
ML Research Engineer at Callosum designing experiments to understand why agentic AI systems fail, building evaluation infrastructure, and publishing findings to improve heterogeneous model/hardware orchestration.
Scale AI · San Francisco, CA; New York, NY
photoroom · Paris
Elsevier · London Wall
Anthropic · San Francisco, CA | New York City, NY
About Us
We’re living through a Cambrian explosion of intelligence: new models and new chips, each specialised for different tasks, are arriving all at once. The result is a new era for AI, one of radical heterogeneity.
Callosum is the Intelligent Systems Company. We believe the next generation of AI won't be defined by any single model or chip, but by intelligent systems in which hardware and intelligence co-evolve. We are building the infrastructure that unifies heterogeneous compute across the full stack. This opens a new axis of scaling intelligence: a dynamic system that tailors itself to what each workload actually needs, whether that's speed, cost, precision, or whatever unit comes next.
The last era scaled on a different bet: one bigger model, more of the same chip, more data. That bet is running into structural limits. Frontier models offer extraordinary capability at unsustainable cost, one that today's monolithic infrastructure was never designed to serve.
Our founding principle is that intelligence comes from many specialised systems working together, not from any single component. We build the software orchestration layer that co-evolves models, workflows and silicon into one system, delivering inference tailored to every workload, and demonstrating orders-of-magnitude leaps in capability and cost.
Because our software spans the full stack, our engineering team works directly with heterogeneous accelerators and frontier silicon, including Cerebras, d-Matrix, Intel, NVIDIA, AMD, Normal Computing, Tenstorrent, GreatSky, and Mixx. We are not stopping at today's chips: each new generation of silicon unlocks algorithms that couldn't run before, and we intend to be first to them, every time. If we get it right, it will belong to everyone building on it - not to any single vendor.
In our latest funding round, we raised $100M, led by Atomico with participation from Plural, DCVC and the UK Sovereign AI Fund’s first investment. With this, we are building the infrastructure for the next era of intelligence.
We are engineers and scientists based in London, working across the full depth of the stack. We are curious, intellectually honest, and building what doesn't exist yet. If you thrive on uncharted territory and are energised by the scale of the challenge, we'd love to hear from you.
About the Role
Callosum believes that orders of magnitude improvements in AI systems will come through application-aware orchestration across heterogeneous models and hardware. Our team is developing the science that makes that possible: a principled, evidence-based method for building agentic systems automatically at unprecedented scale.
Today, designing these systems is a craft. Decisions about memory management, task decomposition, tool use and inter-agent coordination are made on intuition and convention. The results work until they don't, and it is rarely clear in advance which failure comes next.
The team tackles these problems on two fronts. We build the tools to analyse and evaluate agentic systems rigorously enough to say where and why they go wrong, and we use what we learn to design better ones. We are not simply building a better harness; the best harness will change with every new model, task and generation of silicon. We are building the layer beneath it, so that design decisions follow from evidence rather than taste.
The mandate is broad, and most of the questions inside it are still unanswered. You will have wide latitude to choose problems, and your results will shape what the company builds in the future.
What You'll Build
Design and run experiments that isolate why agentic systems fail - across behaviour, traces and activations - and turn those findings into interventions that measurably move intelligence, cost and reliability
Attack one or more of our core research themes: context and memory management; agent steering, task decomposition and specialisation; continual learning during deployment; and inter-agent communication
Build agent analysis and evaluation infrastructure - observability, behaviour analysis, evaluation harnesses, auditing - as durable instruments the whole company works from, not one-off scripts
Publish. Evaluation results, methods and discoveries, as both company assets and public evidence
Work across team boundaries in both directions: take a vertical domain-specific application problem and design systems which run at unpreceding scale at lowest cost and latency; or invent novel model and agent architecture which exploit emerging unconventional silicons beyond GPU
What You'll Bring
Evidence that you can run research of your own: you take an open question, design the experiments that settle it, and produce results other people can build on. Where you learned to do that matters less to us than that you can
Deep hands-on experience with LLMs in agentic settings: multi-step tasks, tool use, long horizons, and the specific ways all of it breaks
Real experimental discipline. Stated hypotheses, controlled comparisons, ablations that isolate one variable, honest uncertainty, and the instinct to distinguish a genuine effect from prompt luck or benchmark noise
Strong engineering. You write the code that runs your own experiments and you are comfortable working inside a substantial shared codebase
Strong communication skills. Able to turn research findings into clear, prioritised guidance for the teams who will act on them, and to write them up for a wider audience
What Sets You Apart
Evaluations, benchmarks or agent harnesses you built that other people went on to use
Experience with RL or post-training for long-horizon, multi-step or tool-use tasks - reward design, environment construction, data generation
Depth in multi-agent systems, planning, program synthesis, or retrieval over structured artefacts such as codebases
Deep familiarity with the internals of SGLang, vLLM, or comparable inference serving frameworks - scheduler design, memory management, and execution pipelines
A background in another field that studies systems of interacting heterogeneous components - neuroscience, distributed systems, economics, cognitive science - and the transfer of intuition that comes with it
A published track record in a relevant field - first-author work at venues such as NeurIPS, ICML, ICLR , ACL or MLSys
What We Offer
Competitive Salary, determined by skills and experience
Equity & Ownership
Private healthcare
We offer Visa sponsorship and relocation benefits to hire the best in the world
We work in person at our London office. You'll have the tools, space and setup to do your best work, and if you have specific needs, just tell us
We're committed to building an inclusive workplace where everyone feels welcome, and believe in equal opportunities for all.
simile · San Francisco