Manasij Haldar, AI Software Engineer

Delhi, IN | +91-7003341179 | [email protected] | LinkedIn

EDUCATION

Chandigarh University Punjab, IN
- B.E, Computer Science Engineering, AI & ML
- Graduation Year: Jun 2024

Brance Technologies New Delhi, IN
- AI Software Engineer Intern (Sep 2023 - Present)
- Developed and implemented conversational bots leveraging cutting-edge technologies, including Generative AI, Large Language Models (LLMs), and Retrieval-Augmented Generative (RAG) models.
- Optimized UChat platform workflows by architecting Python servers for AI recommenders and data engineering pipelines.
- Utilized Flask, FastAPI, Gunicorn, and Uvicorn to enhance performance and scalability.
- Employed Cloud providers AWS and GCP, implementing services like EC2, S3, Lambda, Chalice, CloudWatch, Cloud Functions, and BigQuery.
- Implemented version control using Git for systematic recording of various codes and prompts, ensuring streamlined collaboration and code management practices.

Hyperspace Consulting Remote, Hong Kong
- GCP Data Engineer Intern (Apr 2023 - Aug 2023)
- Engaged extensively with Google Cloud Platform, demonstrating proficiency in tools such as Cloud Functions, BigQuery, and Cloud Composer.
- Architected real-time data pipelines for seamless data ingestion into Data Lake using Airflow and Python.
- Utilized version control tools like Git and BitBucket to systematically store codes for various APIs, ensuring traceability and collaboration in the development process.
- Applied strong database skills, working proficiently with MySQL and Google BigQuery for comprehensive tasks such as Data Exploration, Cleaning, Data Engineering, and Transformation.

ELIXR
- An Automated Image dataset generation Pipeline (Jun 2023 - Jan 2024)
- Innovated and patented a groundbreaking pipeline to extract cinematic key frames from diverse video sources, significantly advancing the creation of high-quality image datasets for Computer Vision (CV) model training.
- Applied expertise in Python, AWS Lambda, S3, GCP Cloud Functions, CloudFlare R2, OpenCV, and Deep Learning to design and implement a scalable and efficient framework for key frame extraction.
- Secured a patent (Application Number: 202341085366) for the unique pipeline, highlighting the substantial contribution to the field of computer vision and image dataset generation.

Hyperspace Consulting Remote
- AI-enabled Crowd Analysis System (Jul 2022 - Aug 2023)
- Innovated and successfully patented a groundbreaking pipeline for extracting cinematic key frames from diverse video sources, significantly advancing the creation of high-quality image datasets crucial for Computer Vision (CV) model training.
- Applied expertise in Python, AWS Lambda, S3, GCP Cloud Functions, CloudFlare R2, OpenCV, and Deep Learning to design and implement a scalable and efficient framework for key frame extraction.

AutoLens Chandigarh, India
- Project Leader (Sep 2022 - Dec 2022)
- Guided the development of AutoLens, a real-time Vehicle License Plate Recognition system, showcasing strategic vision by aligning project objectives and creating a clear roadmap.
- Orchestrated the entire project lifecycle, from planning to execution, ensuring the timely development of a custom-trained License Plate Recognition system using YOLOv5.