About Me
I can run backend infrastructure and ship AI features. Right now building a B2C SAAS as a side-hustle, previously worked at at Markaz Technologies (YC W22), cutting API latency 4x and moving their entire asset pipeline off a third-party CDN. On the research side, I'm an IEEE-published author: my work on LiDAR-based plant phenotyping was presented at AICCSA 2025 in Doha. Good at Valorant. Can type at above 120wpm
Bachelor of Engineering, Software Engineering
National University of Science and Technology (NUST)
Relevant Coursework:
Skills & Tools
Backend & Languages
ML & AI
Cloud & DevOps
Web Development
Work Experience
Software Engineer
Markaz Technologies (YC W22)
- •Migrated AWS Lambda functions to GCP Cloud Run and set up a custom CDN, which reduced API latency by 4x.
- •Moved company assets from Alibaba CDN to GCP buckets and built a pipeline to convert images from JPG to WebP which reduced file sizes by 40%.
- •Developed an LLM-powered API to resolve unstructured address data by mapping addresses to their relevant areas.
Founding Engineer
Eligient
- •Leading the development of an MVP that automates legal contract generation and online dispute resolution.
- •Supervising the technical execution and product strategy for an MVP, meeting with industry experts and researching cutting-edge technologies to enhance features.
Undergraduate Researcher
Machine Vision & Intelligent Systems Lab
- •Conducted literature review of state-of-the-art deep learning models in precision agriculture for plant phenotyping.
- •Worked on MK-X8 Octocopter drone integration, including 3D printing components, mounting sensors, and deploying pipelines on NVIDIA Jetson Orin Nano for data collection.
Machine Learning Research Assistant
TUKL R&D Center
- •Trained deep learning models on the NMT scalp EEG dataset to classify between normal and abnormal EEG signals.
- •Built a conversational chatbot from scratch, implementing a sequence-to-sequence (Seq2Seq) architecture with an attention mechanism in TensorFlow.
Research & Publications
One paper made it all the way to an IEEE conference in Doha, here's what it was about.
AgriFormer: Advancing 3D LiDAR-based Biomass Prediction through Hierarchical Feature Learning
Malik, et al.
Proposed a Transformer-based model building upon BioNet for processing 3D LiDAR point clouds to estimate biomass. Evaluated on a public wheat and triticale dataset (Yanco, NSW 2019), achieving a 16.6% reduction in RMSE over BioNet. Presented at the 22nd ACS/IEEE International Conference on Computer Systems and Applications in Doha, Qatar.
Certifications
A few programs I sat down and finished.
Featured Projects
A few things I've built.
LiDAR-Based Plant Phenotyping for Precision Agriculture
Developing an end-to-end pipeline for non-destructive plant phenotyping using 3D point cloud data.
Umeed - Marketplace for Rural Women Empowerment
A comprehensive marketplace platform designed to empower rural women by providing a digital platform for economic participation and skill-sharing in rural areas.