At Wisedocs, we’re building a stellar team, and we’re proud to celebrate our people through our monthly Wisedocs Brain Power series. Today, we’re sitting down with Aryan Dhar, Machine Learning Engineer and one of the key architects behind Wisedocs’ next-generation AI infrastructure. Aryan joined Wisedocs in March 2025, and in just over a year, has left quite a mark on both the technology and his team. A Pearson Scholar, systems thinker, and someone who likes to know exactly what will break before it does, Aryan brings a rare combination of precision and purpose to everything he builds.
Tell us about your career so far. How did you end up at Wisedocs?
“I came to Canada as a Pearson Scholar at the University of Toronto – it's a full-ride awarded to just 37 students globally each year for academic excellence, leadership, and community impact. At U of T, I pursued a B.Sc. in Computer Science and became deeply interested in machine learning and AI, especially how intelligent systems can be applied to real-world, high-friction problems.”
After graduating, Aryan went on to build ML infrastructure at Cerebras and Equinix, gaining hands-on experience in high-performance computing, large-scale systems, and production-grade machine learning. However, it was Wisedocs’ mission that ultimately drew him in.
“What really drew me to Wisedocs was the story behind it. Connor's experience in healthcare administration and later navigating the claims process after a catastrophic accident highlighted just how manual and broken the system can be – especially for people already going through difficult moments. The mission to bring more speed, transparency, and fairness to the claims ecosystem, and to give individuals more autonomy and support, resonated strongly with me.
At the same time, the technical challenges were exactly what I was looking for. The opportunity to work on scalable inference, training high-quality open-source fine-tuned models, and designing an ML platform capable of supporting 100× growth are all deeply engaging problems.”
“Joining Wisedocs felt like a rare chance to combine meaningful impact with the kind of engineering challenges that push you to think differently.” - Aryan Dhar, Machine Learning Engineer
That intersection of human impact and hard engineering problems is a theme that runs through everything Aryan does, and it’s the same spirit captured in Wisedocs’ CEO Connor Atchison’s end-of-year reflections in A Year of Momentum, where he wrote about building platforms that help claims professionals make confident, defensible decisions. Aryan is one of the people making that possible under the hood.
Give us a day in the life of a Machine Learning Engineer: What’s your 9-5?
“I like to start my day by benchmarking our ML platform and working closely with operations teams to ensure cases are being processed smoothly. From there, I shift into current project work – drafting architectural designs, running experiments, writing code, testing deployments. I usually wrap up by coordinating rollout strategies with customer success and operations, then monitoring performance after hours.”
No two days look the same, which is exactly how Aryan likes it.
You played a central role in rebuilding Wisedocs’ ML pipeline to handle 100x the scale. That's a massive engineering challenge. Walk us through what that actually involved.
"Rebuilding the pipeline was less about a single change and more about rethinking the system end to end. We moved to a Kubernetes-native architecture, introduced Temporal for orchestration, Ray Serve for model serving, redesigned the system to be highly modular so each component could scale independently, and leveraged Generative AI for numerous optimizations."
The results speak for themselves: startup and inference latency dropped by up to 70%, infrastructure costs were cut in half, and the system now supports claims document processing at a scale that simply wasn’t possible before. Those improvements power the recently launched Intelligence Engine – the foundation of Wisedocs’ Claims Decision Intelligence platform.
What was it like building that next generation of AI infrastructure of Wisedocs’ Claims Decision Intelligence, and what does it unlock that wasn't possible before?
“Being part of building Wisedocs’ next-generation Intelligence Engine has been an incredible experience. It’s rare to get to work on the core AI infrastructure that powers an entire industry workflow from the ground up.” - Aryan Dhar, Machine Learning Engineer
“Rebuilding the engine meant not just 100x-ing speed and scale, but rethinking how claims processing automation could actually unlock trusted, auditable insights for insurance, healthcare, and legal teams.
With the Intelligence Engine, we’re enabling things that simply weren’t possible before:
- WisePrep turns chaotic documents into a case-ready workspace with AI medical record summaries, automated medical record chronologies, and deduplication, allowing first-touch review to be 60–80% faster.
- WiseInsights surfaces risk proactively, detecting litigation flags, treatment outliers, and claim inconsistencies up to 30% earlier.
- WiseShare makes secure collaboration seamless, replacing unsecured email with full chain-of-custody audit trails and encrypted document exchange.
Under the hood, the Intelligence Engine is built for enterprise reliability at scale: processing tens of millions of data points monthly, handling massive claim files quickly, and combining modular AI with expert-in-the-loop verification for defensible, audit-ready outputs. This launch isn’t just a product upgrade; it’s the foundation for the future of intelligent, trustworthy claims automation.”
As part of his role, the hardest problem Aryan runs into is anticipating failure before it happens.
“Every time you scale a system by an order of magnitude, entirely new constraints appear. Scaling isn't just about adding compute – it's about anticipating failure points and designing systems that can gracefully handle them.These strategies allowed us to build a pipeline that is resilient, efficient, and capable of handling 100× the workload of our previous system.”
It’s the kind of forward-thinking engineering mindset that fits naturally into a workplace that has earned the Great Place to Work® certification four years running and is recognized among the Best Workplaces for Startups; an acknowledgment that the people doing this work, and the culture they do it in, genuinely matter.
What advice would you give someone looking to break into AI and ML?
“The field is incredibly broad, so it’s important to build both breadth and depth over time. With today’s AI tools, it’s easier than ever to explore new technologies, so my biggest advice is to focus on building end-to-end ML systems: don’t just train models, but take them all the way through deployment.
Understanding how systems work end to end makes a huge difference in how effectively you can iterate and scale. As you build, try to anticipate which components will break first and think proactively about how to address those limitations. Just as importantly, always be clear on why you’re building something; aligning your work with real customer and business needs is what ultimately makes your work impactful."
It’s advice Aryan clearly lives by, and at Wisedocs, he’s found a place where the why is never in question.
Join our scaling team
Here at Wisedocs, we’re not just building a company; we’re fostering a community where talent thrives, and aspirations are realized. Join us in shaping the future of AI claims processing automation solutions and unlocking limitless potential together by exploring career opportunities with Wisedocs.
.png)