Aravindh Kuppusamy

ML Engineer

https://talent.gravityer.com/aravindh-kuppusamy

Key Strengths

Extensive experience in Machine Learning Engineering, particularly in Computer Vision and Deep Learning.
Strong background in MLOps, including designing and architecting cloud-native platforms for data lifecycle, training infrastructure (Kubernetes), and model deployment (AWS ECS, Ray, Redis).
Proven ability to lead and architect complex ML systems, such as multi-modal perception engines for autonomous vehicles and generative AI pipelines for synthetic data generation.
Proficiency in key ML frameworks like PyTorch and TensorFlow, with practical experience in optimization techniques (quantization, pruning, TensorRT, TorchScript/ONNX) for edge deployment.
Experience with Generative AI (Diffusion models, PEFT) and RAG systems (Bedrock, LangChain, pgvector).
Demonstrated ability to work with large-scale sensor data and distributed processing.
Strong project portfolio showcasing practical application of ML algorithms (e.g., object detection, classification, semantic segmentation, reinforcement learning, search algorithms).

Cultural & Operational Fit

Cultural Fit Analysis

The candidate's career trajectory shows a clear focus on Machine Learning and AI, aligning well with an ML Engineer role. The diversity of projects, from academic research to industry applications in autonomous vehicles and robotics, demonstrates adaptability and a broad interest in the field. The experience at startups (Aonics, Gatik) and established tech companies (iRobot) suggests comfort in various organizational structures. The technical leadership roles indicate a proactive and influential approach, which could be a strong cultural fit for teams seeking senior contributors.

Soft Skills & Operational Fit

The candidate's experience descriptions highlight leadership in architecture and R&D, suggesting strong problem-solving, innovation, and strategic thinking. The MLOps and system design work indicates a structured approach to operational challenges and a focus on scalability and efficiency. The detailed project descriptions imply good communication of technical concepts.

AI is analyzing your overall score…

Identifying your key strengths…

Evaluating your skill match against the job requirements…

Assessing your cultural and operational fit

About

I build perception and ML systems for robots that operate in the real world. Over the past decade, I've shipped production ML across consumer robotics (iRobot), autonomous trucking (Gatik), and solar inspection (TerraWise). My work spans the full stack: training and deploying vision models on edge hardware, building MLOps pipelines on AWS/GCP, and designing the infrastructure that keeps robots seeing and deciding reliably in the field. More recently, I've been deep in the LLM space — RAG systems, multi agent architectures, GenAI pipelines for synthetic data generation and model evaluation. Currently exploring agentic AI for robotic manipulation.

Top Skills

Generative AiDeep LearningStable DiffusionPyTorchPythonSensor FusionSystem ArchitectureTechnical LeadershipTensorrtServerlessFastapiKubernetesTensorFlowMachine LearningComputer VisionMicrosoft FabricAlgorithmsData StructuresC++JavaSQL

Education

Rochester Institute of Technology

Master’s Degree, Computer Science

January 1, 2016 – December 1, 2018

Anna University Chennai

Bachelor of Engineering (BE), Electrical, Electronics and Communications Engineering

August 1, 2009 – May 1, 2013

Experience

Aonics

Founding Machine Learning Engineer

October 1, 2025 – Present

San Francisco Bay Area · Hybrid

Gatik

Senior Machine Learning Engineer

September 1, 2024 – April 1, 2025

Mountain View, CA · On-site

iRobot

Senior Software Engineer, Machine Learning Research

August 1, 2022 – May 1, 2024

iRobot

Software Engineer - Machine Learning

February 1, 2019 – August 1, 2022

Rochester Institute of Technology

Graduate Research Assistant

January 1, 2018 – August 1, 2018

Rochester, New York, United States · On-site

Kodak Alaris

Deep Learning Software Engineer

May 1, 2017 – December 1, 2017

Rochester, New York, United States · On-site

Rochester Institute of Technology

Graduate Teaching Assistant

August 1, 2016 – May 1, 2017

Rochester, New York Area

Computer Sciences Corporation (CSC)

Software Engineer

August 1, 2013 – December 1, 2015

Chennai, India · On-site

Projects

Object Detection in Aerial Images

June 1, 2018 – Present

Working on a modification of Mask-RCNN model that can predict oriented bounding boxes in aerial images. Also analyzing the performance of other object detection models like Faster-RCNN, YOLOv3, SSD, and R-FCN in the aerial datasets. TensorFlow, PyTorch. Python.

Visual Question Answering

January 1, 2018 – May 1, 2018

Defined VQA as a multi-label classification problem to predict semantic concept vectors and analyzed whether this training regime decreases language rigidity and make systems generalize better. Made use of CNNs, LSTMs, Attention and NLP techniques. TensorFlow, PyTorch. Python.

Metal Parts Classifier Application

November 1, 2016 – Present

An application to classify different metal parts based on their features using Multi-Layer Perceptron with sigmoid activation function and Decision Trees with Chi-Squared pruning. Written from scratch without using libraries. Technologies used: Python.

Mood Classification of Songs

November 1, 2016 – Present

Based on the features extracted from the song’s metadata, developed classification models to predict song moods by assigning labels to clustered instances of songs. Technologies Used: Python, MySQL, Weka.

A* Search on Rolling Die Maze

September 1, 2016 – Present

Designed three admissible and consistent heuristics to use for the A* search where a constrained die will roll along its edges to find the shortest path in any maze filled with obstacles. Technologies Used: Python.

Tic-Tac-Toe Game

August 1, 2016 – Present

Designed and implemented Tic-Tac-Toe game between a human and a computer where the computer will predict its best move using Min-Max Algorithm with alpha-beta pruning for efficiency. Technologies used: Python.

Instant Messaging Client

April 1, 2016 – Present

Designed an instant messaging application with MVC and Observer design patterns and implemented it with the help of file systems, TCP/IP and UDP protocols. Technologies Used: Java.

Airport Traffic Control System

March 1, 2016 – Present

Designed a system which controls the landing of flights with accordance to the number of landing strips and gates available at the airport using multi-threading. Technology used:Java.

Key Strengths

Extensive experience in Machine Learning Engineering, particularly in Computer Vision and Deep Learning.
Strong background in MLOps, including designing and architecting cloud-native platforms for data lifecycle, training infrastructure (Kubernetes), and model deployment (AWS ECS, Ray, Redis).
Proven ability to lead and architect complex ML systems, such as multi-modal perception engines for autonomous vehicles and generative AI pipelines for synthetic data generation.
Proficiency in key ML frameworks like PyTorch and TensorFlow, with practical experience in optimization techniques (quantization, pruning, TensorRT, TorchScript/ONNX) for edge deployment.
Experience with Generative AI (Diffusion models, PEFT) and RAG systems (Bedrock, LangChain, pgvector).
Demonstrated ability to work with large-scale sensor data and distributed processing.
Strong project portfolio showcasing practical application of ML algorithms (e.g., object detection, classification, semantic segmentation, reinforcement learning, search algorithms).

Cultural & Operational Fit

Cultural Fit Analysis

Soft Skills & Operational Fit

Aravindh Kuppusamy

Key Strengths

Cultural & Operational Fit

About

Top Skills

Skills

Education

Experience

Projects

Certifications

Key Strengths

Cultural & Operational Fit