I'm Sathyajit, aka @satilog on the web and @satilame on X. Here's some info about me!
-
Teaching Assistant @ University of Regina
Taught and evaluated labs for courses on Data Structures and Algorithms in C++.
(Regina, Canada. 2024 - Present) -
Software Engineer @ Watchdog β Focused on R&D
Built knowledge platforms, implemented RAG models, and increased user interaction by 5x.
(Colombo, Sri Lanka. 2022 - 2023) -
Software Engineer @ Rooster Technology
Developed scalable microservices, real-time user alerts, and increased user engagement by 490%.
(Colombo, Sri Lanka. 2020 - 2022) -
Data Engineer @ Yukon Software
Automated ETL pipelines, built geospatial dashboards using PostgreSQL and PostGIS.
(Colombo, Sri Lanka. 2020 - 2020)
- Building products that solve real-world problems.
- Creating efficient systems with semantic search and vector-based models.
- Learning about reinforcement learning and generative AI.
- Languages:
Python
,JavaScript
,TypeScript
,C++
- Frameworks & Libraries:
React
,Next.js
,Flask
,Node.js
- Cloud & DevOps:
AWS
,GCP
,Kubernetes
,Docker
,Vercel
- Data & Search:
PostgreSQL
,MongoDB
,Redis
,Elasticsearch
,Pinecone
- AI & Machine Learning:
TensorFlow
,PyTorch
,Keras
,Scikit-learn
- Integration & Messaging:
GraphQL
,Kafka
,RabbitMQ
-
π NASA Space Apps Challenge (2024):
Developed an interactive exoplanet explorer using React and three.js, with KNN and LLM-based classifiers for identifying habitable exoplanets.
π₯ Demo -
π McGill FIAM Asset Management Hackathon (2024):
Developed a hybrid rolling forecast model using Temporal Fusion Transformer (TFT) for stock predictions. Implemented pipelines on GCP for data preprocessing. -
π Illumine - AI-Powered Educational Platform:
Engineered an AI-powered curriculum generation platform using Gemini AI, reducing grading time to seconds.
π₯ Demo -
βοΈ Clinical Assistant App - MIMIC-IV Data Analysis and Visualization:
Built a data analysis platform on GCP for the MIMIC-IV clinical dataset, integrating classical ML models and BioGPT. Used BigQuery and Cloud Functions for data pipelines and Docker + Kubernetes for scalability.
π₯ Demo -
π₯ Surveil AI:
Developed a semantic search system for surveillance footage, integrating image captioning, object detection, and hybrid vector models with BM25 and ClipViT embeddings. -
π StoreNav - AI-Driven Store Navigation Optimization:
Fine-tuned an Xception-based model to achieve 96% accuracy in segment classification. Implemented graph-based pathfinding with A* and TSP heuristics, achieving sub-second query optimization.
π₯ Demo
- 2nd Place ($2500) - Co.Hacks 2024 by Co.Labs
- Global Finalist - NASA Space Apps Challenge 2024 π
- Best Use of MongoDB - Web3Apps MLH Hackathon 2024 π
- Peopleβs Choice Award ($1000) - 24 Hour AgTech Startup (Canada Farm Show 2024)
π§ loganathansathyajit@gmail.com
π satilog.io
LinkedIn | GitHub
"Reading maketh a full man; conference a ready man; and writing an exact man; and, therefore, if a man write little, he had need have a great memory; if he confer little, he had need have a present wit; and if he read little, he had need have much cunning, to seem to know that he doth not.
β Francis Bacon ποΈ