Smeet Dedhia
FullStack + AI Developer | ML Engineer | Data Scientist
Seattle, WA, USA
smeetd1502@gmail.comResume - PDF
Projects Portfolio →Domain-wise Projects showcase
Education
psst…

University of Washington
Master of Science (MS), Data Science
2024 - 2026
3.9 GPA
psst…

Manipal Institute of Technology
B.Tech Data Science & Engineering
2020 - 2024
3.9 GPA
Experience

University of Washington - IT
Software Engineer
Mar 2025 - Present
Seattle, WA, USA
- Architected data pipelines on AWS (S3, Athena, Redshift) using Terraform to ingest and process org-wide digital assets (websites, PDFs, documents)
- Automated compliance evaluation with WCAG 2.1 standards by integrating Adobe APIs into the pipeline to generate accessibility metrics and reports for all digital assets.
AWS
Terraform
S3
Athena
Redshift
Adobe APIs

Intuitive
Software Engineer (Co-Op)
Jun 2025 - Mar 2026
Sunnyvale, CA, USA
- Developed a framework with Node.js, Langchain & Azure for 15+ Enterprise Apps, enabling rapid infusion of AI into enterprise workflows, resulting in 65% reduction in support tickets.
- Added Spring Boot interface to expose core Java server APIs, unlocking legacy data AI.
- Ensured reduction in Azure & LLM API costs by using on-prem GPUs for RAG and building an LLM inference service adaptive to demand with Kubernetes.
- Reduced chat latency from 90 to 20 sec by tuning multi-threading for optimal GPU memory utilization & batching, also resulting in 4x boost in RAG throughput.
- Curated domain-specific datasets & constructed evaluation benchmarks to quantify AI performance on retrieval precision, grounding accuracy, hallucination rate, task success etc.
- Built NextJS ChatUI with chat streaming for LLM APIs & Server-Sent-Events to broadcast realtime agent status. Integrated UI analytics libraries to track user engagement data.
Node.js
Langchain
Azure
Spring Boot
Java
Kubernetes
NextJS
React
MCP

Microsoft
Data Science Capstone Project
Sep 2025 - Mar 2026
Seattle, WA, USA
- Sequel2SQL: An agentic LLM + RAG framework for SQL error diagnosis, optimization & correction.
- AST based targeting: Designed AST-based, segment-level parsing and transformation layers to selectively process large queries for memory efficiency and apply targeted optimization strategies.
- 6% improvement over baseline on the BIRD-CRITIC Benchmark for NL to SQL task performance.
GPT-OSS
T-SQL Server
LangChain
DsPy
Azure

ION Group
Software Engineer
Jan 2024 - May 2024
Pune, India
- As part of the Core Architecture Team, developed WebSDK serving Fortune500 clients, enabling consistent UI for 25+ downstream apps, optimized for low latency applications in finance.
- Engineered UI components like Calendar widgets, smart tables & autocomplete-search. using state management, event-driven patterns, dependency injection & migrated legacy UI from JS to TS.
- Orchestrated seamless data flow between the frontend SDK and Java REST APIs, leveraging Kafka and Pub/Sub for high-throughput, event-driven messaging.
- Owned feature lifecycle end-to-end with & unit testing, acceptance tests & Jenkins.
Angular
TypeScript
Java
Kafka
Pub/Sub
Jenkins

Dronaid
AI Systems Engineer
Jul 2022 - Sep 2023
Manipal, India
- Integrated CV models with robotics stack across hardware, embedded & software layers. Improved realtime inference for YOLO model from 5 to 12 FPS by switching from python to C++, tuning model precision, and various GPU Optimizations with Triton, TensorRT, NSight, CUDA.
- Performed system-level profiling to identify bottlenecks across CPU, GPU, Memory, I/O. Troubleshooted end-to-end issues across hardware, embedded & software layers.
- Designed a multi-threaded processing to separate camera capture, preprocessing, inference stages for stable frame rates. Using concurrency primitives in OpenMP & MPI.
Computer Vision
C++
Python
YOLO
Triton
TensorRT
CUDA
OpenMP
Skills
Languages
Python
C++
JavaScript
SQL
Java
PySpark
R
CUDA
Frameworks & Cloud
Angular
Docker
AWS Redshift
AWS Athena
Azure Databricks
Spring
AI & ML
PyTorch
LangChain
LlamaIndex
MCP-A2A
Huggingface
Scikit-learn
Tools & Systems
JAX
DsPy
MLOps
DevOps
Microservices
Distributed Systems
Honors & Awards
Achiever's Scholarship
Recipient with 20% tuition fee waiver during undergraduate studies
1st Runner Up - HackRx by Bajaj-Finserv
National level hackathon, winning INR 50K with PheonixGraph
1st Place - AppLab Hackathon
First place winner with INR 25K prize
1st Place - eSamudaay Hackathon
First place winner with INR 25K prize