In healthcare, scheduling is an operation where a missed time zone conversion or a midnight-crossing appointment can mean a patient doesn't get the care they need.
When our team set out to build a proof of concept for HTLH's next-generation scheduling system, we knew the problem was complex. What we didn't anticipate was just how instructive the process would become. The result: a fully functional prototype delivered in under a single workday and a shift in how we think about building software in the era of artificial intelligence.
The challenge: a scheduling platform that couldn't keep up with patient demand
Healthcare scheduling is one of the fastest-growing segments in health IT. The momentum makes sense: virtual visits now account for roughly 1 in 4 outpatient encounters across large US systems, and coordinating care across time zones has become both a competitive necessity and a patient safety issue.
That's the context in which HTLH came to us. Timezone errors and rigid scheduling logic were actively blocking patients from accessing care. We built a telehealth scheduling engine designed to handle the edge cases that break conventional platforms, proved by correctly converting a Chicago Friday 10 PM shift to New York Saturday 2 AM, the exact failure mode HTLH had been living with.
The client's existing platform was built on pre-built components optimised for rapid MVP delivery. That served its purpose early on, but as clinical operations scaled, the limitations became impossible to ignore: time zone conversion errors for availability blocks crossing midnight, fixed 15-minute increments preventing flexible buffer management, recurring appointments booked one by one, and no way for providers to connect their personal calendars.
On the analytics side, leadership had limited visibility into organisational performance. The executive dashboard offered only ten fixed metrics, new reports required manual developer input, and key operational patterns were going unnoticed.
What we built: a scheduling engine for the edge cases that break conventional platforms
Our proof of concept directly handled these gaps. We built a scheduling engine designed to handle the complexity that breaks conventional systems: cross-time-zone scheduling with precision, flexible buffer management, recurring appointment booking, and intelligent availability management that prevents the logic errors that undermine trust in the existing platform.
The moment that validated the entire approach came when the time zone conversion finally worked correctly. When Chicago CST correctly displayed as New York EST, and Friday 10 PM gracefully became Saturday 2 AM — it represented a meaningful technical milestone, confirming that the underlying logic was sound and the architecture was performing as intended.
Where AI helped — and where human expertise made the difference
AI excelled at tasks that consume much time of developers: database migrations, code structure, and boilerplate logic. These are areas where AI delivers genuine, measurable speed gains. Where it struggled was with the nuanced, domain-specific business logic at the heart of this problem.
What surprised us was watching AI stumble over the exact same pain points that plague the client's technical team — time zone conversion complexity and the notorious midnight-crossing problem. It was a reminder that even the best AI tools needed experienced developers to guide them through nuanced domain logic.
Even with a detailed implementation plan in place, the development team had to guide the AI through the solution step by step — identifying where it was going off course and redirecting it toward the correct path. This is not a criticism of AI. It is a clarification of its role.
AI excelled at database migrations and code structure — tasks that would normally consume hours of developer time. However, when it came to the complex business logic of timezone conversions and midnight-crossing transitions, even a detailed implementation plan wasn't enough. The development team had to guide AI through each step, identifying where it went off course and redirecting it toward the correct solution.
The strategic lesson: AI accelerates development, but developers drive the outcome
The 10x speed gains are real, but they are contingent on developers who can recognise when AI has reached its limits, diagnose the problem quickly, and steer toward the solution. The earlier a team identifies where AI is struggling, the faster they can course-correct, and the better the outcome.
This PoC demonstrates that solutions can be delivered 10x faster with AI — but only when developers actively collaborate with it, rather than simply delegating to it. When AI encounters a barrier, the developer's ability to quickly identify the problem and guide AI toward the solution becomes the true multiplier effect.
For business and technology leaders, the strategic implication is clear: investing in AI-assisted development is a sound decision, but the return on that investment is directly proportional to the quality of human oversight guiding it. The right model is not AI replacing developers. It is developers becoming more productive by knowing when and how to intervene.
Ultimately, this is about developers becoming force multipliers: knowing when to step in, redirecting AI when needed, and maintaining the judgment that no model can replicate. The earlier that friction is spotted, the faster the correction, and the stronger the outcome.
The road ahead: from validated prototype to production
The HTLH Scheduling PoC is not just a technical prototype. It is a validated approach to a genuinely difficult problem. It demonstrates that the scheduling complexity inherent to US healthcare can be solved, and that self-service analytics can put meaningful insights in the hands of every stakeholder. The development timelines that once seemed fixed are now compressible, with significant implications for how organisations plan and invest in software delivery. HTLH Scheduling was proof that AI dramatically accelerates development, but complex problems still demand human expertise to chart the course.
The path from proof of concept to production will require continued collaboration among clinical stakeholders, technology teams, and AI tools. But the foundation is solid, the approach is proven, and the potential to improve both provider satisfaction and patient experience is substantial.
AI is your turbocharger, not your autopilot. The magic happens when developers recognise AI's limits early and step in to guide. This is where 10x speed becomes reality.
Related Insights
The breadth of knowledge and understanding that ELEKS has within its walls allows us to leverage that expertise to make superior deliverables for our customers. When you work with ELEKS, you are working with the top 1% of the aptitude and engineering excellence of the whole country.
Right from the start, we really liked ELEKS’ commitment and engagement. They came to us with their best people to try to understand our context, our business idea, and developed the first prototype with us. They were very professional and very customer oriented. I think, without ELEKS it probably would not have been possible to have such a successful product in such a short period of time.
ELEKS has been involved in the development of a number of our consumer-facing websites and mobile applications that allow our customers to easily track their shipments, get the information they need as well as stay in touch with us. We’ve appreciated the level of ELEKS’ expertise, responsiveness and attention to details.
