Development

The ins and outs of AI at Osedea

Rahat Yasir

4

min read

Osedea is an innovative, human-centric, Montreal-based software development firm that builds cutting-edge custom digital solutions to optimize processes and create new revenue streams for their customers. Since the company’s inception in 2011, Osedea has honed its skills in AI-driven technologies in Manufacturing, Automation, and Construction - leveraging a unique combination of applied research and a team of skilled product development engineers. Now, Osedea is capturing the attention of global players. Osedea is strategically positioned to expertly serve both the US and Canadian markets in the Manufacturing, Automation, and Construction industries.

Osedea AI - our go-to market strategy

We complete the AI loop by operationalizing AI models and rapidly putting them into production - setting you up with a new revenue stream or highly optimized operations, fast.

Just like Osedea’s other IT-based offerings, we invite our clients to experience a complete and versatile AI application development process, which includes the following components:

An AI Auditing week
An AI Proof of Concept to define ROI
Production-ready AI integrated application development

AI Auditing week

We believe in empowering customers by sharing knowledge, breaking down requirements into manageable chunks, and giving practical estimates.

Osedea’s AI team collaborates with the client’s team to understand their process and review their infrastructure, AI plan, objectives, and requirements. We also conduct a thorough review of the client’s data to find out what is possible to achieve, and what limitations we may encounter.

The AI Auditing Week is based on an interactive approach that helps our clients feel confident about their investment in our services by defining the ROI they want to see and validating their hypothesis.

AI Auditing Timeline

Day 1: A two-hour interactive workshop on AI business cases, the AI application development life cycle, our AI integration approach, and AI market trends - along with Q&A.

Days 2 and 3: Client shares requirements & sample data, Q&A to understand objectives, data definition, discussion of data format/quantity, available infrastructure, and end goal.

Day 4: Osedea’s team evaluates sample data, reviews the suggested approach, and breaks down the overall requirements with proper justifications.

Day 5: Osedea presents an overall end-to-end solution that is technically feasible and cost-effective. Limitations, proposed quick wins, and different phases of the development process are discussed.

AI Proof of Concept to identify return on investment

According to a recent survey by IDC (International Data Corporation), 65% of the companies studied reported a 10-49% failure rate on their AI projects, while 30% of the companies studied reported a 90% success rate on their AI projects (ref -1). 96% of the failed AI projects ran into problems with data quality, data labelling, and building model confidence (ref-2).

At Osedea, we believe in showing value by building Proof of Concept on smaller scopes to gain the confidence of our clients. Success of custom AI applications is largely dependent on the objective, available pre-trained models, amount of training data, training and inferencing environment, realistic expectations, and iterative milestones. When clients want to design custom AI applications in the domain of machine learning, computer vision, data analysis, or natural language processing, but don't have a way of validating their data and goals, Osedea’s AI team works with them to build a Proof of Concept on their data - with a production mindset. Our PoC approach is to develop the initial proof of concept to identify ROI and verify the hypothesis in a development phase that consists of four two-week sprints.

During the AI PoC phase, the AI developer, cloud developer, software developer, and web and mobile developers work together based on the project needs to design a complete end-to-end product.

Production-ready AI integrated application development

AI enables, empowers, and connects all sorts of digital devices together. Osedea’s AI team works with multiple clients who integrate AI/Data Analytics-based prediction, forecasting, pattern analysis, graphical pattern visualization, anomaly detection, and clustering and flagging unwanted occurrences in their existing web, mobile, and desktop applications. The Osedea team collaborates with the client’s IT team to design AI/Data Analytics capabilities for the existing system. The Osedea AI team examines and interprets the data pattern, acquires the data, and trains different types of explainable and ethically balanced AI models for them. We also deploy them, either on the client’s infrastructure, or as an API in Osedea’s Azure AI instances. Client apps can easily connect to this system using a RESTful API, adding AI capabilities to existing web/mobile/desktop systems.

We bring AI research into tangible reality with an enterprise-graded, production-ready mindset from Day One. Whereas many North American companies working in AI development are focused on research (much of which never leads to the creation of a viable product), Osedea’s goal is to approach building AI-based solutions so that they will inevitably be completely usable - as quickly as possible. This is ensured through engaging in parallel research and software product development phases. The result? The AI loop is completed, bringing an impressive return on investment to Osedea’s customers, with minimal delay.

Production-ready, scalable cloud infrastructure is the pillar of AI application development. Instead of refactoring and rewriting everything before putting it into production, Osedea’s intelligent AI development strategy enables our customer to go live whenever they want to make AI usable.