Leveraging AI and Historical Data for Strategic HR Decision-Making
In today's dynamic business landscape, Human Resources (HR) leaders are constantly challenged with making strategic decisions that can significantly impact the organisation's performance. Traditionally, these decisions have been based on intuition, experience, and limited data analysis. However, with the advent of artificial intelligence (AI) and advanced data analytics, HR professionals now have access to powerful tools that can help them identify the best strategic scenarios in the short term and make informed decisions based on historical data and predictive modelling.
One of the key ways in which AI can assist HR leaders is by analysing data produced from spreadsheet inputs to identify optimal HR strategies. By leveraging machine learning algorithms, AI models can analyse vast amounts of data, including financial, operational, and environmental factors, to identify patterns and trends that human analysts may overlook. This data can include information such as revenue, costs, staff performance metrics, employee engagement levels, and more.
Contextually labelled data plays a crucial role in this process. By labelling data with relevant contextual information, such as the specific circumstances under which certain decisions were made or events occurred, AI models can better understand the context in which decisions were made and identify patterns that lead to successful outcomes. For example, if a particular HR strategy was implemented during a period of economic downturn and resulted in improved employee retention, the AI model can learn from this historical data and recommend similar strategies during similar economic conditions in the future.
Furthermore, AI models can learn from comparing forecasted data to actual figures over time. By continuously analysing historical data and comparing it to real-world outcomes, AI models can refine their predictions and improve their accuracy over time. This process, known as machine learning, allows AI models to adapt to changing circumstances and make more accurate predictions based on past experiences.
One of the advantages of using AI in HR decision-making is its ability to quickly adapt to changing circumstances. For example, if there is a sudden change in market conditions or regulatory requirements, HR leaders can adjust their budget and forecast figures accordingly, and the AI model can realign its analysis to incorporate these changes. This flexibility allows HR leaders to make more agile and data-driven decisions in response to changing business conditions.
In addition to providing data-driven insights, AI can also generate narrative-based scenarios based on past data and learned experiences. By analysing historical data and identifying successful strategies from the past, AI models can generate narrative prompts that provide HR leaders with recommended courses of action based on evidence from past data, rather than relying on intuition or guesswork.
Example of AI for Strategic Analysis using ChatGExample of AI for Strategic Analysis using ChatGPT
First, it must be said that ChatGPT has been known to “hallucinate” and it is important in an enterprise environment to verify the result of AI calculations before the result can be used to lay a platform for future iterations and contribute to machine learning. The next section illustrates how that can be done.
For the purpose of demonstration, I am using an example from 2017 when Competitive Edge Technology introduced a gamification exercise for a Hackathon in India. The exercise was based on a fictious airline and a spreadsheet was used during the preparation stage to verify the custom built Salesforce.com application calculations. The same principle applies here: Trust, but verify. ?
In an enterprise setting a strategic planning process would be accompanied by textual information and would involve carefully labelled data, providing context to the HR issues addressed. Unlike the sophisticated generative AI and predictive AI prevalent today, this method used in 2017 would have been categorised as conventional AI, relying on programmed rules and structured data. However, with the advancement of AI, unstructured data has become integral to decision-making processes.
In this gamification scenario, participants would forecast the financial implications of various strategies, while the program would compute the impacts and determine the optimal solution. The provided example illustrates a snapshot of this approach, showcasing a spreadsheet used for online participants. At the outset lies the Budget/Forecast, commonly used for operational planning across industries. Below, detailed breakdowns illustrate the financial repercussions of each scenario.
Spreadsheet Verification of AI Calculations
The spreadsheet, serving as a prompt, was uploaded to ChatGPT, soliciting an identification of the best-case scenario. The subsequent unedited response from ChatGPT, presented beneath the visual aids, provides insights into the financial outcomes of each strategy and the AI calculations correspond to the results used in the 2017 gamification exercise.
The Unedited ChatGPT Response to the Spreadsheet Upload
The same spreadsheet was uploaded to ChatGPT to test the calculation capabilities of ChatGPT. It did not give the right result first time but after identifying where the discrepancy was occurring ChatGPT recalculated and came up with the expected result, shown below.
Note, the spreadsheet was uploaded as a Microsoft Excel Worksheet (.xlsx) document and not a CSV format to retain formulae.
Scenario RA 1(a):
Scenario RA 1(b):
Scenario RA 1(c):
领英推荐
Narrative:
RA 1(a): Despite the increase in staff costs, the company experiences a significant loss due to the combination of higher expenses and unchanged revenue. Working overtime to compensate for lost time doesn't seem to have a significant positive impact on profitability.
RA 1(b): The increase in passenger revenue is not enough to offset the rise in other costs, resulting in a loss, although it's less than in RA 1(a). Advising staff about termination due to continued absence may have contributed to reduced staff costs, but the overall effect on profitability remains negative.
RA 1(c): Similar to RA 1(b), there's an increase in passenger revenue, but the loss persists due to other costs, resulting in a significant negative impact on profitability. Despite efforts to identify instances of unscheduled absenteeism and counsel individuals, the company still faces challenges in controlling costs.
Conclusion:
All scenarios result in losses, but the best scenario is determined by the magnitude of the loss:
1.??? RA 1(c): -$334 million Despite efforts to address absenteeism and counsel staff, the company still faces a substantial loss, albeit smaller than in other scenarios.
2.??? RA 1(b): -$433 million Advising staff about termination has a slightly more negative impact on profitability compared to RA 1(c), resulting in a larger loss.
3.??? RA 1(a): -$779 million Working overtime to compensate for lost time leads to the largest loss among the three scenarios, indicating that this approach is the least effective in mitigating the impact of unscheduled absenteeism.
Based on these findings, RA 1(c) appears to be the best scenario as it incurs the smallest loss among the three options.
In practice, all of the calculations could be hidden from the person requesting the advice and only the narrative could be programmed to appear.
Another prompt to ChatGPT could ask for more details in tabular format, illustrated below.
Of course, what if scenarios varying the baseline figures can be undertaken, once you are confident the initial calculations are correct. For example, the Passenger Revenue figures could be down graded to reflex a drop in passenger bookings resulting in a different HR scenario being selected as the best option.
Future Implications
AI and advanced analytics hold promise in transforming HR decision-making. By leveraging historical data, contextual information, and machine learning, HR leaders can discern the best strategic pathways in the short term, leveraging past experiences to inform future choices.
In the future, a Chief HR Officer confronted with operational disruptions due to unscheduled absenteeism might swiftly consult an AI application for guidance. Drawing from statistical evidence and machine learning, the response might advocate for pinpointing instances, engaging supervisors, and offering individualised counselling—a testament to the evolving landscape of HR management.
NOTE: Important Comment
ChatGPT has limitations and whilst I used one instance of Risk Avoidance there were five other categories in the 2017 game, each with three HR issues and three scenarios. To process that amount of data requires a larger foundation AI model.
In this example ChatGPT is not creating a permanent knowledge base or data store. ChatGPT only maintains a conversation thread for the duration of the interaction session, allowing users to reference previous messages and context within that session. However, once the session ends, ChatGPT does not retain memory of the conversation or any data shared during it. Each new session starts fresh, without access to previous queries or information. Therefore, ChatGPT does not store or access data from past interactions beyond the current session.
Also, the example was for a fictitious airline. Company confidential operational data does not belong on a public AI model. A private AI model should be used where access can be controlled and data uploaded is intended for company use only.
As we move further into the digital age, the landscape of enterprise systems is undergoing a significant transformation. Traditional centralised systems are gradually giving way to decentralised applications running on Web3 platforms, leveraging the power of blockchain, Distributed Ledger Technology and smart contracts. This shift offers unparalleled transparency, security, and efficiency, disrupting conventional models of data management and operation. It may also influence the positioning of AI components across networks and bring custom built versus services equations into planning.
In light of this evolution, businesses may find it prudent to refrain from overcommitting resources and investment to current AI platforms whose compatibility with the emerging infrastructure of the next decade remains uncertain. Instead, leveraging free, versatile tools like ChatGPT for prototyping and experimentation can offer valuable insights into the potential applications of AI within the evolving technological landscape. By embracing a flexible and exploratory approach, enterprises can adapt more seamlessly to the paradigm shifts expected in the coming years, ensuring that their investments align with the future needs and opportunities of the digital era.
For more information please contact me at [email protected] or https://www.dhirubhai.net/in/john-macy-blockchain-dlt-hr-microservices-solarspacepower-web3platform/?