about us
About US

Who we are?

We are passionate about transforming the fruits of state-of-the-art research and innovation to real-world use cases, which deliver measurable business value. Our expertise is a well-knit puzzle of machine learning, data integration and management and service-oriented enterprise IT architecture and development.

Our portfolio spans across several business domains, including banking, finance, agriculture, retail and law enforcement.

Our Values

Our core belief is that the future of artificial and synthetic intelligence relies on two stepping stones: hybrid learning (HL) and hybrid solving (HS). Both concepts are part of the field of Artificial Collaborative Intelligence (ACI), which enable human and machine to dynamically pass control over a process, throughout its stages depending on automatically tracked robustness, accuracy, precision and efficiency metrics. 

 

The Knosis innovation relies on applying this principle both for solving and cross-validating tasks (HS), but also for data discovery, training on new tools, conversing in foreign languages or practising new skills (HL).

Our mission

A marketplace for machine learning dataset, curated by humans.

 

Our primary objective is human-to-machine (H2M)  learning: delivering high-fidelity, high-depth augmented data, to be used for training, boosting and validating artificial intelligence.  However, throughout our projects and assignments we realized that Knosis helps workers transition into the digital economy by making their skills visible, relevant and valued by a new business landscape, powered by augmented intelligence.

 

As the boundary between human and machine intelligence is blurring, financial compensation is no longer the only expectation of modern digital knowledge workers. They expect that the skills they learn and develop to have long-term, appreciating value. It is for this reason that Knosis uses patent-pending technology to blend-in machine-to-human (M2H) learning, allowing workers and collaborators to acquire new skills regarding to automation, basic scripting and regular expression, in a visual, mobile-friendly environment .

prosthetic-arm-on-blue-background-3913025 (1)
Group-570w
Our Philosophy

The core tenant of our philosophy is that fusing machine and human intelligence will pave the way to efficiencies which will transform all industries. We believe in world where there is no longer a gap of access to automation. 

As far as our technical philosophy is concerned, in our iterative book “Embodied Computational Calculus” (notes open sourced here) we describe in great detail the organization of intelligence around the principle of entropy, negentropy, memory and energy.

Is your business ready for AI?
Our founders
bogdan-portait-bkg

Bogdan Bocse

CEO
_DSC9790-Edit-2

Radu Jinga

CTO
_DSC9717-Edit-2

Alexandru Irimia

Team Lead: SSP
_DSC9553-Edit-2

Andrei Stoicescu

Team Lead: Integration
demo-attachment-55-Path-1631
Group-570w
success Stories
jeshoots-com-fzOITuS1DIQ-unsplash

Document Understanding and Semantic Transcription

Documents, invoices, business cards, letters are not just about the letters they contain. They are about the meaning that they are expected to convey within a specific context, of either business or communication. Based on innovations in latent space fusion, Knosis starts from automatically transcribing the text, while it learns, from how user queries and interactions, what is the meaning of each element, where should it be mapped or looked-up and how it changed throughout time.

Building multi-media dictionaries for machine learning

As the machine learning market matures, the competition shifts from designing capable algorithms to designing efficient algorithms, in terms of quality-for-cost. For this reason, it is paramount to establish a good practice for data capture, augmentation, validation and quality assurance before you decide to invest the time and effort of your development team in pursuing pre-2020 methods of achieving seamless integration between systems.

 

Based on the DeepVISS specification, built using the OpenAPI standard and available open-source, Knosis lets you crowdsource the data tagging for both visual and textual data sets with predictable fulfilment, quality and cost. Your machine learning team should be able to focus on exploring use cases and matching them with better approaches and algorithms, not on reworking code as requirements change. 

Visualizing High Dimensional Data

With knowledge workers having to deal with increasing degrees of complexity, one critical capability is representing such relationships graphically, thus leveraging the ability of our visual cortex to participate in the analysis. 

 

For analyzing complex dependencies between objects of various types, the Knosis tagging console supports high-dimensional visualization, both on desktop and mobile browsers. By blending fuzzy data visualization techniques, such as t-SNE, with logical, categorical and projective techniques, such as the Gestalt principles, we are able to offer business users a fusion between flexibility, intuition, on the one hand, and precision, accuracy, specificity on the other.

Hybrid Conversational Agents

While GPT-3 and other question-answering models are redefining the level of autonomy for conversational agents, our approach as Knosis is to keep the humans in the loop. By this manner, the machine conversational agents continually learn new cases and new interpretations from expert human agents, while at the same time offloading the burden of such menial cases.

 

One significant differentiator of chatbots designed by Knosis is that our model-ensembles include, at the same time, lexicography, graphology, phonetic and semantic dimensions of dictionaries, thus allowing for more depth and breadth for machine intuition.

Bias detection and fair-moderation for data sets

The most pervasive forms of bias are those which we are not aware of. It is for this reason that we recommend to customers with exigent requirements for data quality and algorithms robustness to have their data sets augmented in several folds (independently formed analysis). This allows for the elicitation of hidden biases latent in either the data or in the domain model.

 

By blending a series of lexicographic, semantic, sensory and geometrical measures, Knosis is able to measure both data tag confidence and user bias in regards to a wide variety of data sets, ranging from business documents and diagrams up to satellite and medical imagery.

Begin your journey here

demo-attachment-126-project
0
Project Completed
demo-attachment-129-smile
0
Happy Clients
demo-attachment-125-planning
0
Business Plan
demo-attachment-133-tea-cup
0
Cup Of Tea
Excellence Record

But I must explain to you how all this mistaken idea of denouncing pleasure and praising pain was born and I will give you a complete account of the system, and expound the actual teachings of the great explorer of the truth, the master-builder of human happiness. No one rejects, dislikes, or avoids pleasure itself, because it is pleasure, but because those who do not know how to pursue pleasure rationally encounter consequences that are extremely painful. Nor again is there anyone who loves or pursues or desires to obtain pain of itself, because it is pain, but because occasionally circumstances occur in which toil and pain can procure exercise, except to obtain some advantage from it? But who has any right know how to pursue pleasure to find fault.

Clients