"Passionate about harnessing the power of AI to drive innovation and solve real-world problems."
I am a Machine Learning enthusiast with a keen interest in Generative modelling applications. I have a master's degree in Machine Learning and Computer Vision, and I am currently working at a Bioscience pharmaceutical company in Singapore as an Applied AI Engineer for novel drug discovery for Cancer and Autoimmune disease treatments.
This work made me be able to master working on an HPC, Linux environment, get more proficient in Python/PyTorch, and be able to fine tune Large models for our Applied applications.
Using Generative models for denovo drug synthesis and discovery using models like Diffusion based models and LLMs. During my tenure in Singapore, I delved deeper into Bioinformatics and Microbiology, merging that knowledge with advanced AI/ML techniques to pioneer a revolutionary approach to drug discovery. By leveraging generative models such as RFDiffusion and large language models (ESM3, ProBert), I worked on designing novel antibodies to combat cancer and autoimmune diseases. This synergy between biology and AI has the potential to accelerate breakthroughs and bring transformative treatments to patients worldwide.
Developed robust predictive models (Python, R) to enhance sales forecasting by 30%. Performed data cleaning with SQL & Pandas and advanced EDA with NumPy & Matplotlib, driving data-informed business decisions.
Engineered an AWS-based Agricultural Disease Detection solution using a custom augmented dataset, achieving 90%+ accuracy. Automated image analysis pipelines to serve a global client base.
Led a high-volume operation at one of London's busiest Marugame Udon branches. Oversaw daily team management and cultivated a deep understanding of Japanese hospitality and culture.
The Fashion industry is one of the old traditional industries that have shown massive amounts of reluctance to adapt to the changing tides brought by AI. I decided to help reduce the burdens of fashion designers by creating a tool using stable diffusion models to generate any number of garments from just a text prompt or a sketch.
You can read the full research paper here.
You can use the model here.
Machine Learning for Visual Data Analytics
Electrical Engineering
Email: abhisteak@gmail.com
LinkedIn: LinkedIn Profile
GitHub: GitHub Profile
Download Resume: Download here