What we do

In the Sitroom, we “automate knowledge”; we develop tools to unlock the deepest, seemingly hidden, and most granular understanding – in the form of clear and targeted actions, recommendations, trends, patterns, and insights in general – of how to underpin new business models, enable new roles, and, ultimately, overturn performance. Knowledge gaps – coming from the idiosyncratic functioning of a particular environment, system, team, operation, or facility – otherwise unfeasible to produce at scale by human judgement, are systematically and automatically delivered to fuel new action-taking, new ways of operating.

 

Our challenge

A tool or process that periodically tells store employees when best to refill the cooler, clean the snacks area, place the distributor’s cargo, or double resources at the counter can help adjust activity and skills, speed, and quality of service. A CRM system indicating medical sales reps the actions to take on the following customer visit in order to achieve the planned targets. Or a virtual scheduler telling a truck when to deliver the cargo to a particular client in order to maximize load.

These are typical examples we deal with, where the real challenge consists in finding out how to affect the key constraining paradigms – fixed delivery loads, fixed job schedules, sales reps managing orders – that prevent many businesses from taking off.

 

How we work

We build knowledge automation applications by combining the constant measurement of what people are doing in the context of a particular task or job, usually involving distributed environments such as sales or retail networks, with learning algorithms that identify information gaps to fill and correct. New knowledge is therefore developed iteratively to constantly deliver the right insight at the right time.

We integrate these applications with other systems or technologies – CRM, VR, sensors, image recognition – as needed to build custom solutions. Our core areas of focus are digital twins, insight interfaces, and AI virtual assistants.

Digital twins

We use digital twins, or virtual reality models, to break down the inner workings of a process or business and understand alternative behaviours when a certain number of variables is affected. We apply this to contexts traditionally unfamiliar with these methodologies, such as crowd management, labor management, fleet management, complex project management. Any area or system where several intertwined variables might come into play to produce different outcomes – it can be a retail store, a customer contact center, a supply chain, a business function such as procurement, or a whole company itself – can be subject to granular measurement and modelling, and therefore be a source of input to analytics, AI, and decision-making processes in general.

Insight interfaces

As opposed to the operational representation of digital twins, we here represent the logic of a business; where do margins or growth come from, what operational drivers – i.e., number of customer interactions, number of deliveries, number of product categories, size of staff, route distances, etc. – can be activated to affect each performance variable. This sets the ground on top of which different analytics models can be built to provide insights on any selected variable, especially those that offer unexpected explanations to a particular pattern or behavior, along with attractive visualization interfaces for their regular digestion

AI virtual assistants

Based on the set of operational drivers and performance variables previously modelled, both through digital twins or insight engines, we design different adaptations of learning algorithms, where an agent learns from the experience based on the feedback of its own actions, to iteratively find the best way to control and influence the surrounding environment – customer interactions, supplier negotiations, one-to-one product marketing, etc. – of a process or business. We do this to transform supply chains, understand how to provide better treatment to patients, offer personalized content to customer, augment capabilities of full teams, and many more.
 

Related stories

From Tetris to Puzzles

A view to change the game for transport fleets

Keep reading

Digital labor management

Becoming more efficient while improving customer service (...)

Drones that harvest

Knowing crop yields in real time. Increasing sustainability (...)

Cognitive Procurement

To what extent can Procurement add value beyond cost efficiencies? (...)

Simulated realities

Truths and myths of data intelligence to avoid talking about technology (...)