Machine Learning Application Development

We help you to analyze, learn, process and adapt to the exponentially growing and constantly changing streams of information relevant to your business or research area. Our solutions are built with the KnowmAPI, a collection of code modules built on top of a Thermodynamic Random Access Memory (kT-RAM) emulator. When kT-RAM is available, your application will be able to take advantage of the ground-breaking performance gains offered by this new form of computing substrate.

There are 4 steps from the beginning of a contract or collaboration to the final deliverable.

Step 1: Free Consultation

If you have ideas about what you are looking for, let us know. Otherwise, talk to us us about how you use and gather data. We work hard to understand your specific challenges and will evaluate and suggest possibilities. If there is mutual interest, we proceed to step 2.

Step 2: Free Evaluation

Under strict confidentiality, we will analyze your objective and/or data to determine the feasibility, scope and impact of the project. When we are confident the project can be attained within an agreeable budget, we proceed to step 3.

Step 3: Product or Service Development

Using our machine learning tools developed in the Knowm Development Community, we develop or support your application quickly and efficiently, providing progress updates as we go.

Step 4: Integration

Solutions can be delivered as deployable code, on a dedicated server with the service pre-installed, integrated onto your existing servers or programs, or we can host it for you.

Are you in need of some intelligent automation? Are you drowning in data and need a way to sift through the mess? Do you want to be alerted when something unusual is occurring in your business? Do you have an existing machine learning solution but want to upgrade to the next generation of computing? We can help. We have either already solved the problem and can show you the way, or we can build you a custom solution.
We welcome the opportunity to work collaboratively and in partnership with others, including universities and not-for-profit entities, large and small businesses and individuals to advance intelligent machine technology. We offer group and private workshops for development partners in AHaH Computing and the KnowmAPI. Not a development partner? Talk to us about your needs or ideas for collaborations. Once we have an NDA and joint IP agreements in place, we'll open your eyes to a whole new way of computing.
Through our partner company, Alexander Nugent Consulting, we can act as prime or sub-contractor on US government funded programs. If you have an idea that pairs your technology or skills with ours and would like to pursue a joint solicitation in collaboration with us, let us know.


    • giulio d'agostino


      do you plan to release free workshops videos / material on AHaH Computing and the KnowmAPI in the near future?

      I am a neuroscience student at Trinity College Dublin, studying alternatives to VLSI or PN architecture for emulating neural networks.

      Thank you,

      • Alex Nugent
        Alex Nugent


        Thank you for your interest in AHaH Computing. Tutorials, access to the KnowmAPI and other repositories is available to KDC members. We accept bounties in place of the annual fee, but it is not free. Please check the FAQ page for links to public resources. We are available to answer some questions on our Reddit forum.

    • Dr.K.Paramasivam

      Dear Sir/Madam
      I am basically a researcher and working as Professor in Electronics & Communication Engineering Department, Kumaraguru College of Technology, Coimbatore, TN, India 641049.
      I am working in the memristor based memory cell and memory unit. The work concentrates on identifying better widowing function that characterizes the memristor performance. windowing functions such as stukov, Joklekar, biolek, prodromakis and hybrid window functions are used to analyse the performance. suitable memristor is used to model CMOS-memristor based memory cell.
      In this case, can i use your memristor( BS-AF-CR/W/SN 16DIP ) for designing memory cell? this is my doubt.
      Please clarify.

    • Tim Molter
      Tim Molter

      We’ve developed a new stochastic model for our devices, which also works well for other types of devices: We haven’t specifically tested our devices yet for memory applications as we are focusing on neuromorphics at the moment.

Leave a Comment