Home

I have been coding my whole life

Hi, I'm Hemza. Since I was 13, I've been fascinated by coding and problem-solving, always eager to explore how AI can be used to address real-world challenges. I thrive on tackling complex problems and turning them into scalable, efficient solutions.

For me, technology is more than just improving efficiency, it's about shaping the future. I believe that AI holds the key to solving some of the world's most pressing challenges, from eliminating inefficiencies to unlocking new opportunities. My goal is to be part of the AI revolution that propels humanity into an era of abundance, where intelligent systems empower people to reach their fullest potential and drive meaningful progress.

I'm always looking for like-minded collaborators and curious minds to join me on this journey. Let's connect, exchange ideas, and explore how we can contribute to the future of AI. Feel free to reach out, or subscribe to my updates as I continue to share insights, projects, and aspirations along the way.

About Me

I thrive in fast-paced environments where adaptability and problem-solving are key. Having worked at both large corporations and startups, I've gained a deep understanding of how to navigate different work cultures and drive impactful results.

My passion lies in artificial intelligence. I believe AI has the potential to transform the world and usher in an era of abundance. My ultimate goal is to play a key role in shaping AI-driven technology that not only enhances efficiency but also improves lives, unlocks human potential, and contributes to a future where innovation benefits everyone. I want to be part of the movement that ensures AI is developed responsibly, ethically, and with the power to uplift humanity.

Beyond work, I love rock climbing, staying up to date with the latest advancements in technology, and spending time with family. I deeply value hard work, equality, and honesty, and I bring these principles into everything I do.

Projects Overview

IsenGuard AI

IsenGuard AI is a state-of-the-art security and monitoring system powered by advanced artificial intelligence, designed to elevate both digital and physical safety through intelligent automation. By harnessing the latest in AI and cloud technologies, IsenGuard AI delivers real-time threat detection, automated response capabilities, and smooth integration with enterprise security infrastructures. Due to the success of our project we were selected for an interview with Y Combinator for the Fall 2024 Batch. The code could be seen on the following GitHub link: https://github.com/HemzaZaman/IsenguardAI

Optimx AI

Optimx AI is an innovative project management tool designed to streamline the task management process by automating the creation of Jira tickets for project managers. By monitoring communication channels such as Slack and Zoom transcripts, OptimxAI identifies actionable items and asynchronously generates Jira tickets, significantly enhancing efficiency and productivity for teams. The tool has garnered substantial interest, securing a coveted interview with YCombinator, and attracting over 780 demo sign-ups. Moreover, it has successfully onboarded 150+ project managers from prestigious companies, including Microsoft, for its initial trial. Due to the success of our project we were selected for an interview with Y Combinator for the Winter 2024 Batch. Discover the potential of OptimxAI by viewing the demo video here and explore our website here for more information.Also feel free to look at out our GitHub: https://github.com/HemzaZaman/OptimxAI

Athena AI

Athena AI, our project submission for TreeHacks 2024, marks a significant milestone as we celebrate the 10th anniversary of TreeHacks, the annual hackathon hosted by Stanford University. Inspired by the common student dilemma of finding time to watch lengthy educational videos, AthenaAI introduces Athena and Minerva, two revolutionary AI-powered tools designed to maximize the utility of educational content. AthenaAI leverages advanced technologies to provide a centralized AI-powered repository for educational media, enabling users to efficiently navigate through vast amounts of information and focus on the essential parts of their learning materials. Athena, developed as a full-stack application, and Minerva, a powerful FastAPI backend, work in tandem to facilitate intelligent document querying and topic-specific searches within videos. Our solution not only enhances productivity but also empowers users to manage, explore, and engage with their content in a time-efficient manner. With a robust tech stack including Next.js, Convex, Bun, Langchain, and OpenAI GPT-3.5 Turbo, AthenaAI represents a leap forward in educational technology, offering a glimpse into the future of personalized learning and content management. The report could be seen on the following GitHub link: https://github.com/HemzaZaman/AthenaAI

Forex

Currency conversion fees are usually considered to be unavoidable. But due to the amount of money exchanged globally between different currencies, traders could easily exchange the money directly between themselves, thus avoiding the conversion fees. This website is where people would post what currencies they have and which ones they desire. Once a match is found they meet up and trade. The code could be seen on the following GitHub link: https://github.com/HemzaZaman/Forex

Citadel DataOpen Report

The East-Coast Dataopen hosted by Citadel and Correlation One is an invite-only Datathon featuring university students across the east coast. The event started on the 21st of March and ended on the 28th, lasting about a whole week. I was in a team of two where we analyzed the non-conventional indicators and instigators of traffic. We posed the following question: "How would investments in businesses and education affect traffic and road safety in major American cities?" In this report, we analyzed datasets and later created recommendations for which non-conventional areas municipal governments should invest in to reduce traffic congestion. The three cities of focus were New York, NY, Austin, TX, and Washington, DC. We analyzed publicly available data on 311 calls, building permits, business licenses, population education level, and traffic congestion statistics. The report could be seen on the following GitHub link: https://github.com/HemzaZaman/Citadel-Datathon

OpenCare

In many parts of the world, healthcare inaccessibility is a significant problem, especially for marginalized communities such as low-income individuals, ethnic minorities, and those living in rural areas. This app allows for accessible healthcare based on an online Q&A for medical advice and is a valuable tool for people who have limited access to healthcare services. This app could provide users with a platform to ask medical questions and receive answers from qualified healthcare professionals. The app has a user-friendly interface that makes it easy for users to navigate and find the information they need. To use the app, users input their medical questions, and the app would match them with qualified healthcare professionals who are best equipped to provide an accurate answer. The healthcare professionals could include doctors, nurses, and other medical professionals. The Q&A feature could also include a search function that allows users to look up previously asked questions and answers. This could be particularly helpful for people who have common medical conditions and want to learn more about managing their symptoms. In addition to the Q&A feature, this app also has an upvote and downvote feature allowing expert professionals to be able to critique others' opinions allowing for a more targeted response to ensure the best type of care possible is provided to the patient. This app could help bridge the gap between patients and medical professionals and provide users with the information they need to manage their health. The report could be seen on the following GitHub link: https://github.com/HemzaZaman/OpenCare

Amazon Customer Reviews Scraper

This web scraping project would download amazon reviews using scrapy and its built-in files. To choose which product to download go into reviews.py in spiders and change the values in the list of ASIN numbers to the products you would like to download. Then run the program in the command prompt. The code for the project could be seen on the following GitHub link: https://github.com/HemzaZaman/Amazon-Reviews-Scraper

Vaccinator

I had built a game based on Space Invaders with a Covid-19 spin on it. It was a solo project submitted to Hack-cade, a hackathon hosted by MLH. The project was made using python with the following modules: pygame, os, time, and random. The code could be seen on the following GitHub link: https://github.com/HemzaZaman/Vaccinator

Blog

Thoughts on AI, Technology, and Innovation

Welcome to my blog, where I share insights on AI, technology, problem-solving, and the future of innovation. Here, you'll find my thoughts on the latest trends, personal experiences, and reflections on how AI is shaping the world and leading us to an era of abundance.

Check out my latest posts:

Stay tuned for more updates! If you have any thoughts or topics you'd like me to explore, feel free to reach out.

Citadel Data Open

Python, Plotly, Seaborn, Matplotlib, Pandas, Pandasql

  • Analyzed the impact of investments on traffic congestion and posed solutions predicted to lead to 20.47% reduction
  • Cleaned, organized, and structured multiple data sets with 1 Million+ entries with Python and Dask Dataframes
  • Generated 12+ interactive graphs using Plotly, Matplotlib, and Seaborn while utilizing Pandas for data organization
  • Co-authored a report which examined how investments in businesses and education affect traffic in major cities
  • Applied analysis techniques to derive statistically significant findings on traffic congestion patterns for 3 major cities

Athena

Convex, Bun, FastAPI, Langchain, Next.js, React, shadcn/ui, Clerk, TypeScript, Tailwind CSS, OpenAI GPT-3.5 Turbo, Together.ai, Vercel, Zod, YouTube Transcript Library

  • Developed Athena, a full-stack Next.js application, and Minerva, a FastAPI backend, empowering users to manage and query video and document content. Athena uses Convex for vectorized search, and Minerva leverages Langchain and OpenAI GPT-3.5 to provide fast, topic-specific search functionalities
  • Facilitated multimodal vectorized searches across over 50 educational video transcripts using Convex's optimized database queries, increasing information accessibility by 60%. Incorporated shadcn/ui for seamless user experience and real-time interactions with animations and drop-in UI components
  • Implemented FastAPI and Langchain in Minerva to perform fast, accurate topic-specific searches within videos, achieving response times of under 200 milliseconds. Achieved a 91% satisfaction rate in post-hackathon feedback for improving productivity in educational video consumption
  • Utilized Next.js and React for a responsive frontend and server-side rendering, reducing average page load time by 43%. Integrated Clerk for seamless OAuth user authentication, enabling users to sign in with Google and other providers with minimal friction
  • Obtained 95% positive feedback from participants on the intuitive design and user interaction provided by Tailwind CSS and Zod for form validation, enhancing the overall UX/UI

Forex

MongoDB, Express, React, Node, Firebase, Tailwind CSS, JWT

  • Developed authentication system facilitating up to 300 concurrent logins with an Express.js server and MongoDB
  • Incorporated JWT to maintain authorized user's access to the API and chat reducing average required logins by 78%
  • Utilized Fixer.io's API with optimized consumption to display exchange rates for 170+ currencies as rate suggestions
  • Received a 94% user satisfaction rate in product surveys for the MERN app's facilitation in P2P Currency Exchange
  • Featured a Firebase chat with response latency under 100 milliseconds using React with Gmail for profile creation
  • Obtained a 100% in user interviews for UX/UI of front-end by using frameworks like Tailwind CSS with PostCSS

ATS Corporation

  • Reduced user-reported data latency of values in a new dashboard by 61\% using WebSockets with SignalR in .NET and C#
  • Improved UX by lowering document upload time by 63% through parallelizing many API requests using RxJS observables
  • Optimized database queries, resulting in a 16% decrease in display table load time by decomposing MSSQL commands
  • Created a new dashboard resulting in a 54% UI improvement based on user feedback using Angular with JavaScript
  • Rectified a time conversion bug in the REST API controller, yielding a 14% reduction in testing scripts errors
  • Achieved a 14% reduction in testing script errors by fixing a time conversion bug in the REST API controller of AuthSvc
  • Fixed installer bugs with NSIS related to security and permissions issues, resulting in a 12% reduction in vulnerability alerts
  • Cleaned up code and implemented best practices resulting in a 38% decrease in code smell issues on SonarQube

Achievers Inc

  • Lowered approval time for anonymous posts by 96% with image processing from Google Cloud Vision with Django (Python)
  • Decreased average order email confirmation time by 37 minutes by implementing a real-time data pipeline using PubSub
  • Reduced server load by lowering API calls by 23% by converting endpoints from REST APIs to GraphQL with StepZen
  • Lowered anonymous post approval time by 96% with an image processing Django API with Vision AI for content oversight
  • Updated chat system, reducing response latency by 34% by using GraphQL subscriptions with Redis PubSub
  • Reduced database query response times by 41% by fine-tuning MySQL queries embedded in the Python APIs
  • Created Jenkins CI/CD pipelines with Docker containers and Kubernetes clusters, accelerating prod deployment by 46%
  • Lowered GCP server costs by 30% through containerizing Go microservices and lazy loading them with Kubernetes
  • Lowered profile page load time by 24% by optimizing React components for faster rendering with windowing options
  • Removed external dependencies (e.g. Material-UI) by rewriting components which decreased app load time by 8%
  • Refactored and removed anti-patterns in components eliminating client-side errors by 73% in Sentry error logs
  • Added 1300+ registered users to the platform by enforcing login on the posts page with OAuth 2.0 and JavaScript
  • Eliminated 94\% of accessibility bugs and achieved the WCAG 2.1 standards by adding ARIA labels with indexing

SPARK

  • Developed an authentication system used for over 200 accounts to access quizzes using JavaScript with Node and Express
  • Increased concurrent user capacity by 25% based on Artillery test results through PostgreSQL load balancing with Citus
  • Configured automated emails using Nodemailer, delivering 2K+ emails within 3 min on average confirming quiz results, etc.
  • Cut API downtime by 33% by applying quotas and request throttling in Amazon API Gateway to stop malicious activity
  • Automated the manual user enrollment process using cron jobs with Node-Cron saving 200+ working hours/year
  • Automated the user enrollment process saving 200+ working hours/year with cron jobs using Node-Cron and MongoDB
  • Reduced home page load times by 17% by utilizing CDNs, minifying code, removing unnecessary items and plugins
  • Standardized the design language with a mobile-first approach which improved mobile user satisfaction rates by 29%

CrowdDoing

  • Collected data for 65+ herbs by scraping multiple web sources (ex. NLM) through Python with Scrapy and Beautiful Soup
  • Processed and cleaned 35+ data sources into a unified dataset, reducing data inconsistencies by 98% with Pandas and NumPy
  • Improved query performance by 31% on average by leveraging B-tree indexing and horizontal partitioning in PostgreSQL
  • Developed a recommendation system for herbs/medicinal foods using TF ranking in TensorFlow with ScaNN for retrieval
  • Applied cluster analysis techniques such as K-means clustering to classify 100+ items into their nutrient categories
  • Classified 100+ items into nutrient categories by applying k-means clustering to add filters and improve successive searches

Contact

I'm always open to interesting conversations about AI, technology, and innovation. Whether you have a question, a collaboration idea, or just want to connect, feel free to reach out.

You can contact me via email at HemzaZaman@gmail.com or connect with me on social media. I look forward to hearing from you!