Listing Thumbnail

    Autonomic AI Functor Micro Models - AI Professional Services

     Info
    Autonomic AI professional services provides customized Functor Micro Models for governed, energy-efficient, auditable machine learning in enterprise, streaming, and edge environments.

    Overview

    ` Autonomic AI Functor Micro Models - Professional Services on AWS

    Autonomic AI provides adaptive Functor Micro Models and Small Language Models (SLMs) for enterprise, streaming, and edge environments on AWS. Founded in 2025 by John Harby - a software architect with 35+ years of experience and former Senior Architect for HP e-speak (the first web services implementation, which won a Smithsonian award) - Autonomic AI is a member of the AWS Partner Network. The architecture focuses on one-shot learning and energy-efficient machine learning, designed for regulated and high-scale environments requiring auditability, observability, rollback, lineage tracking, and adaptive intelligence.

    This listing is offered as a professional service because custom setup through the AssuredModel process can significantly optimize your model. To begin an engagement, provide Autonomic AI with: (a) the model type, (b) your training data, and (c) your optimization criteria - such as latency or accuracy. The AssuredModel process then determines the optimal model function selection for your specific domain and requirements.

    Proven Energy Efficiency and Performance

    Functor Micro Models deliver measurable energy reduction compared to conventional AI approaches. Using the RBF (Radial Basis Function) micromodel as a representative benchmark on AWS bare metal with RAPL measurement: Energy per inference: Approximately 30.9 mJ per inference (92,630 J over 3,000,000 inferences) Throughput: Approximately 6,483 predictions per second Accuracy: 89.27% best validation accuracy on a 10-class dataset with 9,469 training samples

    The architecture combines the LLM Offloading Pattern (LOP) with Recoverable Learning to eliminate the two largest energy sinks in modern AI: the quadratic cost of multi-turn agentic conversations and the waste of full-scale stochastic retraining. Recoverable Learning replaces irreversible parameter mutations with governed semantic learning operations, enabling deterministic recovery bounded by the local dependency cone rather than full history replay.

    Horizontal Scalability by Design

    Functor micro models consist only of functions with no mutable state, making them inherently horizontally scalable. This is a fundamental architectural property, not an operational configuration choice. Transparent failover clustering is straightforward because writes can be applied to all members in a commit sequence while reads can be distributed by a load balancer. A live cluster demo is available at the resource links below,

    Highlights

    • Governable, auditable and transparent. The model learns by changing function(s) rather than parameters. Every change to the model can easily be explicitly logged in an event log such as Kinesis or MSK.
    • One-shot learning is naturally supported. This means the model can continually learn and contain current updates.
    • Functor learning offers governance avoiding model corruption. A rollback function is also provided with each event log entry so the model is governable in the sense of capability to rollback to prior versions.

    Details

    Delivery method

    Deployed on AWS

    Pricing

    Custom pricing options

    Pricing is based on your specific requirements and eligibility. To get a custom quote for your needs, request a private offer.

    How can we make this page better?

    Tell us how we can improve this page, or report an issue with this product.
    Tell us how we can improve this page, or report an issue with this product.

    Legal

    Content disclaimer

    Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.

    Support

    Vendor support

    jharby@functormodel.ai  (951) 740-2650 https://functormodel.ai  demos: https://api.functormodel.ai  Available PDT from 9am to 6pm