Software engineer with 6+ years building high-throughput distributed systems and low-latency trading infrastructure at Amazon and Morgan Stanley. Specializing in ML infrastructure, backend systems, and cloud-scale architecture.
Where I've worked
Building systems at scale across cloud, fintech, and ML — with a track record of measurable impact.
What I've built
From LLM infrastructure to systems programming — projects that go deep.
Fine-tuning platform that lets users train any open-source LLM on custom datasets — no ML expertise required. Handles data ingestion, training, and inference in one workflow.
Investigating offloading LLM tokenization to NVIDIA Smart NICs via Triton Inference Server to reduce time-to-first-token. Research conducted at Santa Clara University.
Built a relational database from scratch in C++ — storage engine, SQL query parser, full CRUD, JOIN operations, and B-tree indexing for efficient query execution.
Real-time restaurant discovery app using geospatial hit-counting on lat/long coordinates instead of static star ratings — surfaces trending spots based on actual visit density.
Technical toolkit
Languages, frameworks, and infrastructure I work with.
Academic background
M.S. Computer Science & Engineering · GPA 3.85 / 4.0
Research: LLM tokenization offloading to NVIDIA Smart NICs · ACM · Hack for Humanity · Competitive Programming Club
B.E. Information Technology · University of Pune
First rank in class · Consistently top 20 percentile of students.
What people say
Sudarshan joined my team for the 2024 summer internship and made a significant contribution to the assigned project, working closely with senior R&D and product managers across geographies. He ramped up on the codebase and dev infrastructure within a week, with major functionality working within the first month. He used solid Java, data structures, and algorithm skills to deliver on performance goals. An eager & quick learner, great communicator who strives to go above and beyond. I recommend him highly for software engineering roles.
Sudarshan, my former reportee at Morgan Stanley, made significant contributions to our systems. Within just six months, he became the go-to person for system-related matters. His technical prowess was evident as he reduced latency in the pricing engine, indirectly benefiting the firm's revenue. His achievements led to well-deserved promotions and awards. His exceptional talent lies in his ability to quickly grasp new concepts — both technical and functional — and apply them efficiently without much effort.
I had the pleasure of working with Sudarshan during his 2024 summer internship, and I was thoroughly impressed by his performance. He ramped up quickly, demonstrating a strong technical grasp of Java and an impressive attention to detail. He is an exceptionally fast learner — whenever there was an issue, he acted swiftly and efficiently. I am confident he will be a valuable team player and a tremendous asset to any team he joins.
Sudarshan is a good professional. He is technically very talented and has good communication skills. He is quite flexible and easy to work with. He is a fast learner and a great developer. I highly recommend Sudarshan.
Sudarshan has been a glorified star of our team. He took very little time in understanding our company's organizational values, the working and our culture. In a short period since joining, he was able to completely harmonize with his peers and deliver high-quality releases with incredible production support.
Sudarshan and I worked together on a Hackathon (which we won!) and also on setting up regression pack for integration tests on automation tools. He has brilliant grasping power and understood complex systems with utter ease which demonstrated how capable technically he is — it's been a pleasure working with him.
Sudarshan has been excellent in performing all the tasks. He worked with me on setting up ELK stack for the QA team at CLSA for analyzing logs which reduced manual effort of filtering through logs drastically. He is very strong technically.
Get in touch
Open to discussing distributed systems, ML infrastructure, or any interesting engineering challenge. Recruiters, engineers, and curious minds — feel free to reach out.