The Biological Computing

What The Biological Computing does

The Biological Computing Co. (TBC) is a privately held deep-tech company founded in 2021 by neurosurgeons Alex Ksendzovsky (CEO, MD/PhD) and Jon Pomeraniec (COO, MD/MBA), headquartered in San Francisco's Mission Bay with an additional location in Baltimore, Maryland. TBC operates a biological computing platform that interfaces cultured living neurons — grown on high-density multielectrode arrays of 4,096 electrodes with approximately 100,000 neurons per dish — with conventional AI architectures. Data is encoded into electrical stimulation patterns delivered to the living neural networks, and the resulting spatiotemporal responses are recorded, decoded, and packaged as compact Biological Adapter modules that integrate into state-of-the-art AI models without retraining.

TBC's commercial offerings center on two product lines. The Biological Adapters line includes the Neural Dynamics Adapter (~156k parameters) for generative video stability, the VAE Adapter (~2.36M parameters) for image reconstruction, and additional computer vision adapters — all marketed at less than 1% the size of the models they enhance and delivering claimed gains of ~19% video image quality, ~27% reconstruction quality, ~23x video efficiency, and 3x faster training convergence on CIFAR. The Algorithm Discovery Platform uses biologically derived principles to propose new AI learning rules, including a local plasticity rule reported at ~97% accuracy on continual learning tasks. The company maintains ~25 full-time employees and is supported by advisors including Stanford Professor Shaul Druckmann and former Google executive Richard Holden.

TBC monetizes primarily through licensing Biological Adapters as software add-ons to AI developers and enterprises, supplemented by a planned managed-service offering of hybrid neuro-silicon cloud clusters targeted for 2027. The go-to-market combines enterprise field sales, product-led technical content marketing via a four-part blog series, and design-partner co-development with named partners such as Odyssey (generative video) and unnamed cybersecurity customers. TBC achieved the first-ever paid commercial transaction for biological compute in 2026, and is positioning its roadmap — including World Models, Real-Time Biological Compute, and Pattern Completion capabilities — toward the broader AI infrastructure market.

Name

The Biological Computing

Legal name

The Biological Computing Co.

Website

https://tbc.co

Company type

Private

Founded year

2021

Operating status

Operating

Headcount range

11–50 employees

Short description

The Biological Computing Co. (TBC) is a San Francisco-based deep-tech company that integrates living neurons cultured on multielectrode arrays with AI models to produce Biological Adapter software modules, improving computer vision and generative video performance and reducing compute cost for AI developers and enterprises.

Ownership category

akta.pro rank

Product category

Biological AI Infrastructure Software

NAICS

Software Publishers (5132), Research and Development in Biotechnology (except Nanobiotechnology) (541714)

SIC

Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ)

akta.pro secondary industry

AI/ML-Enabled Biologic Discovery Platforms (computational design, in silico screening) (HLAAAIAE)

Keywords

• Biological computing
• Neural adapters
• AI infrastructure
• Generative video optimization
• Algorithm discovery

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices3 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Biological Adapter Licensing / SaaS: TBC licenses its biologically derived software adapters (Neural Dynamics Adapter, VAE Adapter) to AI developers and enterprises. These adapters are integrated into existing AI workflows as add-on modules. The company previously operated under the name Biological Black Box and launched the world's first biological computing platform for computer vision and generative AI workloads.
2. Biological Compute as a Service: TBC became the first company ever paid by a customer for biological compute, marking the first commercial transaction in the biological compute category. This signals an emerging market for hybrid neuro-silicon cloud services. The company plans to launch hybrid neuro-silicon cloud clusters by 2027 to target the AI infrastructure market.
3. Algorithm Discovery Platform: TBC abstracts neuron behavior into algorithmic discovery products, offering a software adapter or algorithm discovery platform that helps customers define next-generation AI architectures. This is positioned as a long-term product alongside the immediate Biological Adapter offerings.

Go-to-market motion3 records

Distribution channels3 records

Marketing channels6 records

Core offering

The Biological Computing Co. (TBC) develops software adapters and an algorithm discovery platform that apply living neurons cultured on high-density multielectrode arrays to existing AI models. Its Biological Adapters integrate neural-derived representations into computer vision, VAE reconstruction, and generative video models to improve accuracy and stability while reducing compute cost, without requiring full model retraining. The Algorithm Discovery Platform uses biologically derived principles to guide development of next-generation AI architectures beyond transformers.

Product overview

The Biological Computing Co. (TBC) offers a biological computing platform consisting of two core product lines: Biological Adapters and the Algorithm Discovery Platform. Biological Adapters are modular software components that integrate living neurons (cultured on multielectrode arrays with 4,096 electrodes) with AI models to improve performance and reduce compute costs. The Neural Dynamics Adapter specifically targets generative video stability while the VAE Adapter improves image reconstruction quality. The Algorithm Discovery Platform uses biologically derived principles from real neurons to discover new learning algorithms for AI systems. TBC is also developing future products including World Models, Real-Time Biological Compute, and Pattern Completion capabilities, with plans for hybrid neuro-silicon cloud clusters by 2027.

Differentiator

Problem solved

Functional benefit

Brands

• Biological Adapters: Real neurons used to improve performance and reduce cost of foundation models.
• Algorithm Discovery Platform
• Neural Dynamics Adapter

Products and services

• Biological Adapters Software modules derived from living neural network dynamics that integrate into state-of-the-art AI models to improve performance and reduce compute costs without full retraining. The adapters are less than 1% the size of the models they enhance and are designed for AI developers and enterprises building vision and generative AI systems.
• Neural Dynamics Adapter A lightweight, biologically derived software module (approximately 156k parameters) that stabilizes predictions during inference-time sampling in generative video models, enabling longer and more coherent video rollouts. Achieves approximately 19% improvement in image quality and maintains 5% improvement over LoRA-adapted models in long-horizon stability. Demonstrated on the Oasis DiT Minecraft video model.
• Biologically Derived VAE Adapter A lightweight adapter module that interfaces with variational autoencoder intermediate representations during reconstruction, improving image detail and fidelity without retraining the base model. Achieves a 27% improvement in reconstruction quality relative to the baseline VAE, with only approximately 2.36 million additional parameters (~1% of the base autoencoder).
• Algorithm Discovery Platform A research platform that uses biologically derived principles from real neurons to define and discover new algorithms for AI systems, particularly addressing architectures beyond transformers and overcoming plasticity loss in deep networks.

Quantifiable outcome

• 23x improvement in video model efficiency compared to silicon-based approaches
• +9 more outcomes

Named customers3 records

Segments4 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability7 records

Feature5 records

Partnerships

Five partnerships are on record, tiered key advisor, design partner and reference platform.

• Shaul Druckmann (Stanford Professor)key advisorStrategic or Co-development Partner · 27 May 2026Stanford Professor Shaul Druckmann was appointed as TBC's Neuroscience Advisor as the company enters its commercialization phase, bringing academic neuroscience expertise to guide the company's biological computing platform development.
• Richard Holden (Former Google Executive)key advisorStrategic or Co-development Partner · 27 May 2026Former Google executive Richard Holden was appointed as TBC's Product and Partnerships Advisor as the company enters its commercialization phase, providing product strategy and partnership development expertise from experience at Google and other technology companies.
• Odysseydesign partnerStrategic or Co-development Partner · 22 April 2026TBC is actively partnering with Odyssey, a video generation firm, to demonstrate its biological computing technology. This partnership is cited as an example of TBC's go-to-market strategy of engaging design partners in the AI developer ecosystem.
• Unnamed Cybersecurity Companiesdesign partnerStrategic or Co-development Partner · 22 April 2026TBC is working with cybersecurity companies to demonstrate its biological computing technology and real-time adaptive systems capabilities, exploring applications in security-relevant pattern recognition and prediction.
• Oasis (Open-Source Minecraft World Generation Model)reference platformTechnology or Integration · 1 February 2024TBC uses the open-source Oasis 600M-parameter Minecraft world generation model (Diffusion Transformer) as a reference platform to demonstrate its Neural Dynamics Adapter technology. This is not a commercial partnership but an integration with an open-source model for technology demonstration.

Scale indicators10 records

Recent moves8 records

Expansion highlights6 records

Emerging players

• Groq: Groq provides LPU inference hardware optimized for low-latency, high-throughput AI inference — a primary target use case for TBC's adapter-driven compute efficiency. Competes for enterprise AI infrastructure budgets using silicon advantages rather than biological substrates.
• Mythic AI: Mythic develops analog compute-in-memory AI chips for low-power edge AI inference. Pursues biologically inspired, energy-efficient AI compute — paralleling TBC's broader efficiency narrative though with silicon rather than biological implementation.
• GrAI Matter Labs: GrAI Matter Labs (now GrAI) develops neuromorphic-inspired AI accelerator chips using dataflow architectures inspired by biological neural networks. Targets similar low-latency, low-power AI inference workloads as TBC's adapter offerings.
• Cerebras Systems: Cerebras builds wafer-scale AI compute systems targeting large-model training and inference workloads. Competes for the same enterprise AI infrastructure dollar TBC is targeting, though with fundamentally different silicon-based architecture.
• SynSense: SynSense builds ultra-low-power neuromorphic processors (Speck, Xylo) for always-on AI inference targeting edge and sensor applications. Comparable to TBC's brain-inspired compute ambitions, though implemented in mixed-signal silicon rather than biological substrates.
• Lightmatter: Lightmatter builds photonic AI processors targeting data-center AI compute with claimed substantial energy efficiency gains. Competes for the same enterprise AI infrastructure efficiency thesis that TBC advances, albeit with photonic silicon rather than biological substrates.
• SambaNova Systems: SambaNova offers integrated AI hardware-software stacks (DataScale, SN40L) for enterprise training and inference. Targets the same AI infrastructure optimization pain point TBC addresses for enterprises seeking compute cost reduction.

Broad incumbents

• Intel (Neuromorphic / Hala Point): Intel's neuromorphic research program (Loihi / Hala Point) pursues brain-inspired compute at scale within Intel's broader silicon portfolio. Relevant as an incumbent exploring TBC's adjacent paradigm, with substantially greater R&D resources but also broader focus and slower iteration speed.

Direct peers

• FinalSpark: FinalSpark develops biocomputing platforms using brain organoids interfaced with multielectrode arrays — directly paralleling TBC's MEA-based living-neuron approach and biological-compute-as-a-service positioning. Both are early-stage efforts to commercialize neuron-based compute for AI workloads.
• BrainChip Holdings: BrainChip produces neuromorphic AI processors (Akida) inspired by brain neuron spiking dynamics for low-power edge inference. Both TBC and BrainChip pursue brain-inspired AI compute, though TBC uses biological substrates directly while BrainChip uses digital silicon implementations of similar principles.

Market position

Strengths5 records

Weaknesses4 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/biological-black-box-bbb-inc
• Xx.com/BioComputingCo

Revenue estimate

Valuation estimate

Number of profiles

Profiles10 records

Funding overview

Funding rounds4 records

Investors9 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about The Biological Computing