Passionate AI engineer focused on building impactful systems that drive real business value.
I specialize in applied AI, from optimizing smaller LLMs with supervised fine-tuning to designing multi-agent systems and adaptive assistants with memory and reasoning. I’m pursuing my Master’s in Computer Science at Stevens Institute of Technology and have built AI systems that power enterprises and solve real business problems. My work focuses on making advanced AI methods practical and valuable in real-world applications.
I believe AI should help people achieve more — that’s what drives my work.
Shipped AI Projects
Years Experience
Lines of Code
Cups of Coffee
The principles I return to when making tough decisions, solving problems, and shaping my work.
Focus on 3 things that needs to be achieved in the next 18 hours that move your mission forward; everything else is noise.
Take full responsibility for the work. Act with initiative, solve problems proactively, and drive lasting impact as if the entire mission depends on you.
Place the end-user at the core of every decision. Design, build, and measure success based on the real-world impact delivered to them.
Technologies evolve fast. Explore, experiment, and keep your knowledge sharp.
Go beyond just writing code. Build solutions that move the needle, simplify lives, or create measurable value.
Move fast to deliver initial value, then refine for efficiency, reliability, and polish.
Give feedback that Aims to Assist and is Actionable; receive feedback with Appreciation, then thoughtfully Accept or Discard it.
Converting "impossible" prompts into "impressive" demos, one all-nighter at a time.
“Lawgic AI is a voice-based AI that allows users to call in, describe their legal issue, and get instantly matched with the top 3 relevant lawyers. The system collects and summarizes case details, presents them to lawyers for acceptance, and connects the client with the best available match, streamlining the process of finding legal help in real time.”
“CodeGlance is an MCP server powered by Gemini that analyzes GitHub repositories and auto-generates detailed documentation, architecture flow, and file insights — making it easier for developers to quickly understand and navigate unfamiliar codebases”
“Identified a real-world civic problem and developed an AI-powered solution to automate complaint reporting, making the process faster, easier, and more accessible for new yorkers”
Building AI systems and driving innovation across startups and enterprises.
Mastered end-to-end AI product development from concept to production, learning to build autonomous systems that bridge business requirements with technical implementation
From fine-tuned LLMs to multi-agent AI systems and full-stack applications.
Clarity AI turns YouTube’s vast educational content into structured, interactive learning experiences. It uses advanced natural language processing to transform long-form videos into engaging, organized study materials—making passive watching an active way to learn.
Fine-tuned Llama 3 (8B) using PyTorch and 4-bit QLoRA to optimize RAG performance on finance dataset and GPU memory, implementing tokenization and adapter merging to reduce VRAM usage by ~60% and enable high-throughput on a single T4 GPU.
Want to see more of my work?
View All Projects on GitHubIndustry recognition for innovation, performance excellence, and competitive achievements.
Stevens Institute of Technology
2024
SquareYards
Apr 2024
SquareYards
Dec 2023
Industry-recognized credentials demonstrating continuous learning and expertise.
Amazon Web Services
Amazon Web Services
Stanford University / Coursera
Stanford University / Coursera
Stanford University / Coursera
Stanford University / Coursera
Google Cloud
Proficiency across languages, frameworks, AI/ML tools, and cloud technologies.
Current focus on building, optimizing, and deploying intelligent systems
Enhancing performance of compact models through advanced fine-tuning and optimization techniques
Exploring MCPs to create seamless AI-tool integrations and extensible agent architectures
Building AI assistants that remember past conversations and get better over time
Advanced Computer Science studies with focus on AI, machine learning, and data systems.
M.S. in Computer Science
B.E. in Information Technology
Peer-reviewed research in air quality prediction and IoT-based health monitoring systems.
What colleagues and leaders say about working with me.
“I had the pleasure of working with Harshil during his time as an AI Engineer at Square Yards, and I was consistently impressed by his passion for AI/ML and his drive to build impactful, end-to-end AI systems. Harshil has a solid grasp of a wide range of model architectures, and more importantly, he brings a thoughtful approach to applying them in real-world scenarios. What really stood out was his ability to deeply understand a problem—not just from a technical angle, but from a business perspective as well. This balance allowed him to propose solutions that were both innovative and practical. Any team would be lucky to have him—he's the kind of engineer who not only builds models, but also helps shape the vision of what AI can do for the business.”
Open to AI/ML opportunities, collaborations, and innovative project discussions.
I'm always excited to take on new challenges and collaborate with amazing people. Whether you have a project in mind, want to discuss opportunities, or just want to say hi, feel free to reach out!
Loading...
Location
Loading...