Article

Category: Expertise strategy

Using AI to streamline resource allocation

In this article, we explore a case from emagine HQ where successful resource allocation substantially increased productivity. At the heart of it? Machine learning. Keep reading for more AI optimisation examples, and consider the hurdles and pitfalls.

In the day-to-day of business operations, artificial intelligence (AI) is turning time-consuming, manual tasks into automated processes, achieved within seconds. The cumulative impact of this is substantial, with hours of employee time freed up for other tasks, which could increase project manager productivity by as much as 40%.

Businesses are seeking to harness these new technologies en masse, with Statista forecasting that by 2027, global digital transformation spending will reach $3.9 trillion, up from approximately $2.15 trillion in 2023.

In this environment, consultants such as emagine’s experts are ever more in demand to deliver these highly specialised transformation projects, but allocating the best people is labour-intensive.

Read about AI's potential to accelerate software product time to market

Projected spending on digital transformation technologies from 2017 to 2027

Recommendation engine

At emagine, we begin more than 800 new projects every month, each requiring a specific blend of skills. To streamline candidate allocation, we have developed an AI-driven recommendation system that transforms unstructured data from a database of almost 300,000 candidates into high-quality shortlists in seconds. This has significantly streamlined our recruitment process.

Check out our ebook on AI: Dancing with the Machines

Jan Wolff, Chief Operation Officer at emagine, explains: “At emagine, we are harnessing new technologies to enhance our competitiveness in the market. By developing this premium digital tool, we also increase our efficiency in finding the best experts for our clients with the assistance of Machine Learning.

Close to 300,000 consultants are currently in our system, with more candidates added daily. In 2024 so far, 82% of consultants allocated to projects were already in our database, and we are proud that this indicates that our new recommendation engine is having an impact.”

Leveraging machine learning

The model uses machine learning to analyse the CVs of consultants on the firm's extensive database but also reviews historical project data to assess individuals' suitability for particular projects.

Within approximately 5-10 seconds of a request being created, the model can provide a shortlist of relevant candidates that can be entered into an interview process. This saves the team up to 10 minutes for the primary search, equating to around 1,400 hours annually across approximately 8,000 consulting jobs.

What are the pitfalls?

A risk in this scenario is to rely too heavily on AI because there are considerations outside of pure, hard facts. In a business environment increasingly dominated by digital experience, it's important not to lose the human element.

"In 2024 so far, 82% of consultants allocated to projects were already in our database, and we are proud that this indicates that our new recommendation engine is having an impact.”

Jan Wolff, COO at emagine

Jan says: “Despite AI's time-saving aspects, the human factor is vital in the process and should not be underestimated. Only by actually talking with candidates can our recruiters truthfully evaluate their skills, personalities, and experience to see if they match the client's needs. Our colleagues have vast experience in this, which machines cannot yet process.

The time saved allows us to reach out to the best candidates quickly and to spend more time communicating with clients and consultants significantly faster than before.”

Over-reliance on AI can cause issues. The Institute of Student Employers cites an example where an AI-powered recruitment tool that scanned resumés to identify relevant candidates had a significant bias against female candidates.

Read the Institute of Student Employers article

If you only use AI solutions, you only view a candidate's profile in text and may miss the skills that shine through during interaction. Soft skills do not translate as well into black and white as professional credentials.

Another significant pitfall is the data itself.

Jan adds: “It took three years to develop the model we use at emagine and a significant investment of time into cleaning the candidate data we hold so that it functions efficiently. Attempting to implement AI solutions before the data is ready would result in inadequate results. A suitable time investment is essential to ensure the data is clean.”

How else could AI improve workflows?

AI can be a very beneficial addition to a project team, particularly in relation to information management, project planning, budgeting and risk management. Henrik Timm, Business Unit Director and Partner at emagine, explains that AI’s ability to streamline processes, assist in decision making and enhance automated routine tasks makes it increasingly relevant for project managers to implement.

He says: “One of the primary ways AI can assist in project management is its ability to manage vast amounts of data, but not only that, it can ensure this data is well organised so that it can be utilised most effectively. By leaning on technology to support project management systems, it frees up time for PMs to focus on more strategic areas of the project.

“Projects benefit from AI’s ability to analyse vast amounts of data much faster and more accurately than humans, which allows for more informed decisions which can amplify project success. AI also provides an extra layer of confidence by identifying potential risks and recommending mitigative strategies before they impact the project. This proactive approach to risk management can save substantial time and resources, ensuring that projects are completed within scope, on time, and within budget.” 

"By leaning on technology to support project management systems, it frees up time for PMs to focus on more strategic areas of the project."

Henrik Timm, Business Unit Director at emagine

What are the hurdles?

A major hurdle in any AI use is data quality. The data is the backbone of the whole process, so if there are problems with the data, there will be problems with the AI model too.

Integration with existing systems can also block some uses of AI. For AI systems to function, there must be a constant exchange of data, so if there are any issues when integrating AI with existing systems, this can become a major hurdle.

The great benefits of implementing AI can also be lost if you don't have the skilled personnel to manage it. As with any major technology, AI needs constant maintenance, and this requires specialised skills that can be hard to find, especially experts in ML, Neural Networks and Natural Language Processing (NLP).

AI reflection before adoption

Businesses need to be clear on what the desired outcome is and establish if implementing AI solutions is the best way forward. Organisations should identify the pain points or inefficiencies that AI could address and ensure these align with strategic priorities, as AI will require investment.

When creating a roadmap and implementation plan for AI, it needs to be flexible. It’s also important to embrace the new culture that this technology will bring to a company and its employees. Ensuring internal education and upskilling around AI to build that culture is essential for bringing employees on board, and employee buy-in will ensure a better response and the potential to reap more of the technology’s benefits.

The data quality needs to be constantly maintained. Once it’s up and running, AI can always be further optimised, whether it’s improving response times, the quality of a step-by-step interaction, or how it integrates with other systems or models.

Ready to find out more?

Learn how emagine can support your business through major digital transformation.

Blog

See insights