# Kinjal Haldar

> SWE @Plural

Location: New York, New York, United States
Profile: https://flows.cv/kinjal

vibe coder 😎 (jk)

Currently at Plural working on K8s management and some AI stuff
Formerly at Amazon Q (QBusiness) / AWS Kendra (AI Search AWS service) on the Control Plane.
Graduate from University of Washington, Seattle (B.S. Computer Science).
I enjoy working on computationally intensive tasks such as machine learning / computer vision, backend programming, and more.
I have prior experience working on robotics (perception), machine learning, and server side programming.

## Work Experience
### Software Engineer @ Plural
Jan 2025 – Present | New York, United States

### Software Development Engineer @ Amazon Web Services (AWS)
Jan 2022 – Jan 2025 | New York, New York, United States
Amazon Q (QBusiness) / AWS Kendra Control Plane Team
- Developed toolchain to analyze infra usage by cluster and resize / delete clusters based on size and storage utilization
- Used toolchain to downsize large clusters, reducing monthly non-prod infra costs by ~$2 million (25%)
- Helped build Amazon Q's QBusiness service, mostly on controlplane side (APIs, WFs)

### Software Intern @ Directed Machines
Jan 2021 – Jan 2021 | Seattle, Washington, United States
- Improved accuracy of follow-me algorithm on agricultural robotics system to reduce false-positive target acquisition. Implemented logic to accurately check depth of RGB features by correlating depth and color images with different FOVs. Implemented auto-calibration of depth profile dimensions to avoid setting parameters manually. Added tolerances to account for grainy images. Implemented feature into Java microservice without reducing framerate.
- Wrote utilities in Python to capture and convert raw depth and color camera samples and use them in testing.

### Undergraduate Researcher @ University of Washington
Jan 2020 – Jan 2021 | Seattle, Washington, United States
• Working with Professor Gavriel Kohlberg on visual speech recognition program for people with hearing challenges. Currently developing a module for real-time lip detection from video frames, replacing existing Xbox Kinect dependency.
• Wrote a 2-phase pipeline in Python – phase 1 finds face and phase 2 finds lip points on face. Used a pre-trained Tensorflow model for phase 1. Used Tensorflow/Keras to create and train my own CNN model for phase 2, including preprocessing routines for training/test data.

### Undergraduate Researcher, UW Seattle @ University of Washington
Jan 2019 – Jan 2020 | Seattle, Washington, United States
- Working under Professor Shih-Chieh Hsu in Department of Physics on analyzing particle physics data where I integrated the CONTUR statistical analysis tool for model validation into end-to-end workflow. Used Python for the integration.
- Deployed a server on Google Cloud to run TensorFlow models for electron energy reconstruction. Used Kubernetes to create server instances.

### Member, Husky Robotics Team, UW Seattle @ Husky Robotics
Jan 2018 – Jan 2020 | Seattle, Washington, United States
- Wrote a module in C++ to use GPS data and obstacle detector module to execute next step in navigation. Wrote and tested UDP networking code to communicate with base station.
- This year, I wrote LiDAR processing components in C++ for obstacle detection. This includes algorithms for clustering, filtering, boundary approximation, and map localization. Also wrote a visualizer for LiDAR in OpenCV, including visualization of clustering and boundary algorithms.

### Intern @ Directed Machines
Jan 2019 – Jan 2019 | Seattle, WA
- Worked on computer vision tasks for autonomous agricultural robots
- Integrated TensorFlow models into Java microservice stack, including use of transfer learning and use of quantized models for faster inference
- Integrated a LiDAR into camera streaming server using a C api and made an OpenGL renderer to verify functionality
- Wrote Java and C code for depth image filtering, accounting for camera pitch


## Education
### Bachelor's degree in Computer Science
University of Washington
Jan 2018 – Jan 2022


## Contact & Social
- LinkedIn: https://linkedin.com/in/kinjalhaldar

---
Source: https://flows.cv/kinjal
JSON Resume: https://flows.cv/kinjal/resume.json
Last updated: 2026-04-01