Abhishek Rana
| Experience | Education | Projects |
About
I am a passionate software developer, with 8+ years of industry experience, interested in building (micro)services and a background in deep learning and computer vision.
Experience
MAY 2020 - AUGUST 2023, 3 yrs 4 mos
Software Engineer II, Edge Case Research, Munich, Germany
- Worked in a cross-functional agile team to develop software for BMW to validate safety of autonomous vehicles.
- nLoop Virtual Test House (VTH) to create and run safety test cases across different platforms at BMW.
- Platform Evaluation Service to compare Safety Performance Indicators (SPI’s) between different platforms.
- Developed python microservices with RESTful APIs using hexagonal architecture from ground up.
- Migrated CI pipeline from Jenkins to Github Actions and achieved 50% reduction in execution time.
- Implemented automated nightly deployment and end-to-end testing pipeline using Github Large Runners.
Skills: Python, Golang, Vue.js, Javascript, HTML, CSS, Flask, FastAPI, Pydantic, SQLAlchemy, Pytest, Apache Airflow, Docker, SQL, PostgreSQL, NoSQL, Redis, RabbitMQ, InfluxDB, Helm, Kubernetes, Openshift, AWS, Terraform, Github Actions, RESTful APIs, Agile Methodologies.
MAY 2019 - APRIL 2020, 1 yr
Machine Learning Engineer, Terraloupe, Munich, Germany
- Developed deep learning training and inference pipelines for semantic segmentation using TPU’s.
- Developed/trained object detection algorithms/models for detection of buildings, roads, lanes, etc in aerial imagery.
- Developed 3D lane model from aerial imagery semantic segmentation, DTM and DSM.
Skills: Python, Tensorflow, Pytorch, Keras, OpenCV, Shapely, GIS, QGIS, Docker, GCP, Computer Vision, Deep Learning.
JAN 2015 - AUG 2016, 1 yr 9 mos
Senior Software Engineer, Aricent Technologies, Gurgaon, India
- Developed Timing and Synchronization module to provide accurate time to the eNodeB (LTE) using GNSS and to maintain the timing precision.
- Integrated PTP (Precision Time Protocol) framework (Microsemi) to provide sync to multiple PTP Clients (eNodeB) using PTP server.
Skills: C, C++, Python, LTE (4G), PTP.
JULY 2012 - DEC 2014, 2 yrs 6 mos
Software Engineer, Reliance Jio Infocomm Limited, Gurgaon, India
- Developed SBC (Session Border Controller) to set up, conduct and tear down media communications in a LTE network.
- Implemented audio/video call, call hold and call resume functionality in SBC.
Skills: C, C++, SBC, LTE (4G).
Education
OCTOBER 2016 - APRIL 2019
Master of Science (M.Sc.) in Informatics, Technical University of Munich, Munich, Germany
- Thesis: Adversarial Data Augmentation Techniques for Skin Lesion Classifiers.
APRIL 2008 - MAY 2012
Bachelor of Technology (B.Tech) in Computer Science And Engineering, National Institute Of Technology (NIT) , Hamirpur, India
Projects
Full Stack
Tech stack for microservices based application (Github)
- Modern tech stack for microservices with support for local development/deployment on kubernetes.
| Service | Technologies |
|---|---|
| Frontend | Vue3 - Typescript, Pinia |
| Backend | Python - FastAPI, Pydantic, SQLAlchemy (PostgreSQL), Beanie (MongoDB) |
| Golang - Chi | |
| Infrastructure | PostgreSQL, MongoDB, Redis |
| Deployment | Kubernetes (k3d), Helm |
Skills: Python, Golang, Vue.js, Javascript, HTML, CSS, FastAPI, Pydantic, SQLAlchemy, Docker, PostgreSQL, MongoDB, Redis, Helm, Kubernetes.
Backend
Image to text using AWS Serverless (Lambda) and SAM (Github)
- Image to text service using AWS Lambda function and AWS SAM(Serverless Application Model).
Skills: Python, AWS Lambda, AWS Serverless, SAM.
Frontend
Project Manager (Github)
- Application to manage projects within a team.
Skills: Vue, Vuetify, Typescript, Javascript, HTML, CSS, Vite
Tic Tac Toe (Github)
- Tic Tac Toe game
Skills: React, Javascript, HTML, CSS, Vite
Deep Learning & Computer Vision
Apparel detection using deep learning (Github)
- Detect apparels in images using deep learning classification and regression model trained on clothing images.
Skills: Python, Keras, Augmented Reality, Computer Vision, Deep Learning.
Guitar fretboard detection and notes projection (Github)
- Train and predict guitar fretboard using deep learning model (UNet).
- Project a mesh on the detected fretboard (using Homography).
- Display the notes to play on the projected mesh.
Skills: Python, Keras, Augmented Reality, Computer Vision, Deep Learning.
ISIC 2018 - Skin disease classification challenge (Github)
- Train and predict skin cancer using Cascaded DenseNets for Multi-Class Skin Lesion Classification and ISIC 2018 - Skin Lesion Classification for Melanoma Detection dataset.
Skills: Python, Tensorflow, Computer Vision, Deep Learning.
Image generation using Generative Adversarial Networks (GAN) (Github)
- Generate synthetic/fake images using DCGAN (Deep Convolutional Generative Adversarial Networks) and small NORB dataset.
Skills: Python, Pytorch, Computer Vision, Deep Learning.