Human Resources teams have never lacked purpose. They champion people, shape culture, and are central to employee wellbeing. Yet in reality, HR teams can find themselves buried under a myriad of rules, exceptions, data requests, compliance checks, eligibility questions all   Instead of coaching and developing talent, they answer questions like:  “What happens to my benefits if I take a parental leave?” “What does the waiver of premium cover mean?”       This was the challenge the LACOteam behind HAIRspray set out to solve. With SAS technology and support, they built an AI-powered HR assistant capable of understanding employee life events, applying organisational policies, and automatically triggering relevant workflows across systems.   Their mission was ambitious: use AI to return time, trust, and clarity to HR.   “Putting AI into HR — that’s HAIRspray… giving HR back to HR.”   Understanding the Complexity Problem HR teams today are custodians of a complex regulatory, policy and benefits universe., medical coverage changes after life events, eligibility rules tied to tenure and contract type — all constantly changing. When employees get married, have children, take parental leave, or update their employment status, dozens of rules and dependencies are triggered. One of the LACO team summed up the burden: “HR officers are struggling to master the labyrinth of rules and procedures.”   Much of this workload is manual and reactive;policies live in PDFs, email threads, and intranet pages. HR staff rely on memory, sticky notes, and checklists: “How can I get rid of all my post-it notes and reminders?” The team started by articulating this pain. They had on board an SME in the domain of HR, who clearly set out the operational and strategic issues.The team then mapped a candidate set of processes, rules, data , questions and insights that would showcase their vision for HR.   Framing the Vision The team aligned on a bold ambition: to transform HR from reactive administration into proactive, intelligent orchestration. Powered by AI and SAS, the system would recognise employee life events as they happen, apply the correct rules to each unique situation, and launch the necessary steps — all while ensuring HR could review and intervene when needed. Employees and benefit partners would receive timely, tailored communication without the manual burden.   As the pitch put it: “Recognizing data-driven events and matching them with all applicable rules and procedures. It then automatically initiates the necessary actions.”     Building HAIRspray Experience had taught the team to start simply, therefore the team began by focusing on one of the most persistent and time-consuming tasks in HR: tracking unused holiday balances and reminding employees before they expire. Rather than forcing HR staff to manually sift through records and send individual messages, they built an automated flow that did the work for them.   The system built an agent using SAS, that gathered the relevant data, calculated remaining leave days, and drafted personalized emails. HR could review and approve messages before they went out — or, where appropriate, let the system send them autonomously. As one team member explained, the pipeline “picks each employee, identifies how many days remain, creates the subject and body, and — once HR approves — sends it.”   After proving the power of automation in this everyday workflow, the team moved on to a scenario with even greater complexity and impact: family status changes and the resulting shifts in medical coverage. Marriage, birth, and other family events are notoriously rule-heavy, triggering cascading updates across benefits, eligibility systems and documentation.   Their solution recognized these life events in the employee data, interpreted the appropriate policies, and initiated the required actions — including drafting messages not only for the employee, but also for benefit providers. HR could step in when needed, but the system did the heavy lifting, from detecting changes and validating eligibility to preparing insurer-ready records.   Together, these early use cases showed how routine HR processes — once dependent on memory, manual checks, and email reminders — could become proactive, precise, and largely self-managing, freeing teams to focus on people rather than paperwork.   Technical Foundation The system combined structured HR data with policy documents converted into machine-readable formats: “We pulled the benefits-related fields… turned the key tables and policies into readable and citable data sources… built a single knowledge source directly in SAS.”   Key components: The HR agent consisted of A data store providing structured and unstructured realistic HR data A decision engine acting as the brain and orchestrator. An LLM providing the natural language interface and output. AI models supporting the HR agent’s decision process.   The AI layer ensured explainability and trust: “Every answer points to a source. Time is saved, mistakes drop, and trust goes up.”   A Valuable Realisation: Synthetic Data &  HR AI Agents. One major challenge the team discussed is the difficulty inherent in designing and delivering AI is getting data. Real HR datasets contain sensitive personal information. Training AI systems on them requires governance, masking, oversight from different stakeholders. Much of the time the simple act of accessing data can take months.   Synthetic data removes those barriers by creating realistic, statistically accurate HR data without exposing real employee .   Synthetic data enables: Safe LLM training without privacy risk Testing automation for rare HR events (marriage, adoption, disability leave) Robust scenario simulation Faster model iteration without data-access bottlenecks Bias-testing across demographic attributes   This is where SAS Data Maker becomes a strategic accelerator. With it, organisations can: Generate lifelike employee datasets — tenure, benefit elections, family structures, contracts Simulate complex HR populations across industries Create edge-case scenarios to test automation resilience Continuously refresh training data as policies evolve   Imagine AI systems that continuously learn HR rules safely, without touching real identities. Imagine stress-testing benefits workflows using thousands of synthetic families, life events, or contract types before going live.   For projects like HAIRspray, synthetic data turns student prototypes into enterprise-ready products. It enables secure collaboration, scalable experimentation, and auditable development pipelines —   The HAIRspray team delivered more than a clever name. They demonstrated what’s possible when motivated innovators meet the right data, platform, and mentorship. They redesigned HR work from first principles — not as compliance administration, but as a domain where AI can elevate human impact.   With automated event handling, AI-generated workflows, semantic policy understanding, and synthetic data on the horizon, solutions like HAIRspray signal a new era: HR liberated from routine, HR empowered by intelligence and HR returned to humans   My role as a mentor This was, I am happy to report was a simple task. The team, a longstanding partner of SAS, understood the domain, what they wanted to achieve and how SAS technology could deliver on that promise.   We met on Friday’s to go over ideas and challenges the team had.   My input was to provide experience on how SAS’s AI capabilities had been used in other domains.  It is always a pleasure to meet partners and especially those that have a passion for technology and how it can solve real world business situations. I also thought their team name was more than a touch of genius.!   - David Ferguson A huge thank you to our sponsors, Intel and Microsoft, whose support made this Hackathon possible. Their contributions went far beyond sponsorship- bringing expertise, resources, and inspiration that fueled innovation. Together, they created a truly collaborative and transformative experience for all participants.   LACO  is just one of many visionary teams from this year’s Hackathon. Find out who the champions are during our award session Dec. 11 on YouTube and LinkedIn! ... View more