Callosum
Category: AI Infrastructure
System-level software platform that orchestrates AI workloads across heterogeneous chips and models to enable 'Heterogeneous Intelligence' — diverse AI models running on different hardware types working together as an integrated, co-evolved system. Callosum was founded in 2024. The company is led by Danyal Akarca. Based in London, United Kingdom. Team size: 11-50. Total funding raised: $110.25M. Latest round: Seed. Key investors include Atomico (Lead investor - $100M Seed, Aug 2026), Plural (Led $10.25M Pre-Seed Feb 2026; participated in Seed), DCVC (Seed participant), UK Sovereign AI Fund (first-ever equity investment; Seed participant), AlbionVC (Seed participant), ARIA (UK government R&D grant support / Scaling Inference Lab), Charlie Songhurst (Angel), Stan Boland (Angel - FiveAI), John Lazar (Angel - Royal Academy of Engineering).
AMW Analysis
Callosum develops a system-level software platform that orchestrates AI workloads across heterogeneous chips and models, enabling different hardware types—such as Nvidia, AMD, AWS Trainium, Cerebras, and SambaNova—to work together as an integrated system. The company claims its approach delivers 2x accuracy, 7x faster performance, and 4x lower cost compared to homogeneous single-chip solutions, and up to 12x cost reduction on specific workflows. Based in London and founded in 2024, Callosum has raised $10.25M in pre-seed funding from investors including Plural, ARIA, and the UK government’s Sovereign AI fund.
Recent news flow shows significant government backing and a clear strategic direction. In April 2026, the UK Sovereign AI fund made its first £500 million equity investment in Callosum, alongside allocating up to 1 million GPU hours to seven startups. This investment underscores the fund’s focus on breaking the Nvidia-centric compute monopoly and boosting UK AI sovereignty. Prior news in March 2026 highlighted the $10.25M raise and positioned Callosum as part of a growing wave of Nvidia challengers, with the platform targeting enterprises seeking to run AI workloads across diverse accelerator chips. The company’s trajectory points toward expanding infrastructure for multi-chip orchestration, with a neuroscience-inspired approach originating from Cambridge University.
AMW analysis, generated from 5 tracked news signals.
- Founded
- 2024
- Headquarters
- London, United Kingdom
- Team size
- 11-50
- Total funding
- $110.25M
Value proposition
Delivers 2x accuracy, 7x faster performance, and 4x lower cost compared to homogeneous single-chip solutions by exploiting diversity across models and hardware. Achieves up to 12x cost reduction on specific workflows while setting new SOTA on benchmarks using only open-source models.
Products and solutions
Tailored Inference (GA Aug 2026 — family of APIs routing workload blocks to optimal models/chips across heterogeneous accelerators), Heterogeneous Intelligence Platform (core orchestration software), Programmable Heterogeneity / block-based workload decomposition, Heterogeneous Recursion Engine (deep context reasoning across mixed models/chips), Heterogeneous Vision-Language-Action System (web automation and multi-modal agents), Topology-aware Cache Management System, On-die Grammar Enforcement Kernels (AWS Inferentia2 NKI)
Unique value
First company to co-evolve heterogeneous chips and intelligence together, inspired by brain architecture principles — the human brain achieves intelligence through diverse specialized cell types working together, not by copying one neuron billions of times. Vertically integrated 'Intelligent System' where workflows are hardware-aware, models are task-graph-aware, and kernels are output-constraint-aware, with every layer co-optimized in context of the whole.
Target customer
Companies building multi-agent AI systems requiring superior performance across complex workflows; emerging chip manufacturers seeking to demonstrate hardware capabilities at scale; enterprises with heterogeneous AI workloads (automation, inference, complex reasoning)
Industries served
AI Infrastructure, Enterprise AI & Automation, Cloud Computing, AI Chip Manufacturing, Data Centers, Inference-Optimized Compute
Technology advantage
Combines neuroscience-inspired heterogeneous architecture with full-stack optimization (from kernels to workflows). Achieves: 2x accuracy/7x speed/4x cost improvements on complex workflows; 25% improvement over ICLR 2026 SOTA on Visual WebArena using zero frontier API calls; up to 12x cheaper workflows (SambaNova GPT-OSS-120B vs GPT-5); 2.4x speedup via topology-aware caching; 1,767x faster grammar enforcement at batch-64 through on-die masking; multi-cloud/multi-chip support (AWS, GCP, Azure across Nvidia, AMD, Cerebras, SambaNova, Trainium/Inferentia).
How they differentiate
First company to co-evolve heterogeneous chips and intelligence together with brain-inspired architecture. Achieves 2x accuracy, 7x faster performance, and 4x lower cost compared to homogeneous single-chip solutions by orchestrating AI workloads across diverse hardware (Nvidia, AMD, AWS Trainium/Inferentia, Cerebras, SambaNova) with neuroscience-inspired heterogeneous computing principles.
Main competitors
Ray/Anyscale, Nvidia, Cloud providers with custom chips (Google TPU, AWS Trainium/Inferentia)
Key partnerships
Cerebras (flagship partnership Aug 2026 — WSE integration into Tailored Inference for ultra-low-latency heterogeneous agentic inference), AWS (Trainium/Inferentia; first major cloud; NKI kernels), Rebellions (Korea — next-gen silicon partnership), Axelera AI (Netherlands — inference chip co-design), Intel (inference orchestration / datacenter partnership), NVIDIA (NVIDIA Inception; GB300 NVL72 / DGX Spark workload distribution), Supermicro (OEM / multi-cloud heterogeneous deployments), Normal Computing (thermodynamic computing design partner), SambaNova, AMD, d-Matrix, Tenstorrent, Furiosa, Tendrils, Lumai, Mixx, Cortical Labs (silicon ecosystem), ARIA/CommonAI Scaling Inference Lab, UK Sovereign AI Fund (first-ever investment), HelmGuard AI (Tailored Inference production partner — agentic cybersecurity)
Notable customers
HelmGuard AI (Tailored Inference production partner — agentic cybersecurity; 77x cheaper / 10x faster vs frontier alternatives per Callosum), Coworker AI (enterprise production partner for GitHub workflow benchmarks), Financial services agentic workloads (production via Cerebras partnership — 4x faster, 70% lower compute cost, 10% higher task success)
Major milestones
Company founded November 2024 (Companies House #16076949; renamed from SERNN Ltd Aug 2025), Emerged from stealth February 2026 with $10.25M Pre-Seed led by Plural, First-ever investment of UK Sovereign AI Fund (announced April 2026; closed in Seed), Raised $100M Seed led by Atomico with Plural, DCVC, UK Sovereign AI, AlbionVC (Aug 20 2026) — one of Europe's largest seed rounds, Launched Tailored Inference APIs generally available (Aug 2026), Flagship Cerebras partnership for ultra-low-latency heterogeneous agentic inference, Named in UK £1.1B AI hardware plan, Custom NKI kernels for AWS Inferentia2; multi-cloud heterogeneous orchestration
Growth metrics
Raised $100M Seed (Aug 2026) led by Atomico ~6 months after $10.25M Pre-Seed; total disclosed funding ~$110.25M. Tailored Inference GA; LinkedIn company size 11-50 (~41 employees listed). Production: financial-services agentic workloads 4x faster / 70% cheaper / +10% success vs single frontier+GPU; HelmGuard 77x cheaper / 10x faster on cybersecurity inference.
Market positioning
Early-stage UK-based AI infrastructure startup challenging Nvidia's dominance by enabling multi-model, multi-chip AI workloads. Targets two customer segments: companies building multi-agent AI systems requiring superior performance, and emerging chip manufacturers seeking to demonstrate hardware capabilities at scale.
Geographic focus
United Kingdom (London), Europe, with global reach through multi-cloud platform support (AWS, Google Cloud, Microsoft Azure)
Patents and IP
No registered patents publicly disclosed as of February 2026. Core IP resides in proprietary heterogeneous orchestration algorithms, topology-aware runtime, and custom silicon kernels.
About Danyal Akarca
Cambridge-trained computational neuroscientist and medical doctor. PhD candidate at University of Cambridge's MRC Cognition and Brain Sciences Unit, researching generative models of brain networks. Previously Research Fellow at Medicalchain. Co-author of research published in Nature Machine Intelligence on brain-inspired AI systems.
Latest news about Callosum
- Callosum raises $100M seed round led by Atomico for cross-chip AI infrastructure software
- UK Sovereign AI Fund Makes First £500M Investment in Chip Interoperability Firm Callosum.
- The UK’s £500 m Sovereign AI fund has made its first equity investment in Callosum, a heterogeneous AI orchestration platform, and allocated up to 1 million GPU hours to seven startups. Callosum’s sof
- London-based Callosum raised $10.25M to orchestrate AI workloads across heterogeneous chips, challenging NVIDIA's GPU monoculture. Their platform delivers 2x accuracy, 7x faster performance, and 4x lo
- Zettafleet, a stealth AI startup from Cambridge University's Machine Learning Systems Lab, has unveiled its first product targeting Nvidia's dominance in LLM training with claims of 74-75% cost reduct
More AI Infrastructure companies
Official website: https://www.callosum.com