Aerin Kim

The truth will set you free.

<h1>Welcome to <br> Aerin Kim's website! </h1>
<marquee>I like developing user-pleasing products using data.</marquee><br><br>

<div class="content-description">
 &nbsp;&nbsp; I am a passionate, entrepreneurial engineering lead with extensive hands-on execution experience at both small and large firms. I have a common sense approach to ML/engineering problems and an aptitude for improving user experiences. I thrive on generating results.<br></div>

<links style="line-height: 2;">
<a href="https://automata88.medium.com/" target="_blank">Blog</a><br>
<a href="https://github.com/aerinkim" target="_blank">GitHub</a><br>
<a href="https://twitter.com/aerinykim" target="_blank">Twitter</a><br>
<a href="https://www.linkedin.com/in/aerinykim/" target="_blank">LinkedIn</a><br>
<a href="https://stackoverflow.com/users/3907250/aerin" target="_blank">StackOverflow</a>
</links>

&nbsp;&nbsp; My professional work has been centered around building and optimizing large-scale production-grade machine learning models. Areas that I have sunk substantial time into are:
<br><br>

• ML Applications Development: Various families of semantic parsing (text-to-SQL, question-answering, attribute extraction, machine translation, etc.), chatbot, object detection, semantic segmentation, visual QA, multimodal summarization, cross-modal retrieval, deepfake detection/generation as well as targeted ad delivery. 
<br>

• Search Accuracy and Database Integrity: Improve search result fidelity through advanced deduplication and indexing, leading to operational efficiency and cost savings.
<br>

• Stochastic Modeling and Simulation: Use tools from probability theory, including Markov chains, queueing theory, and reliability analysis, to simulate real-life phenomena. 
<br>

• Training Dataset Synthesis: Datasets for academia and industry using both synthetic and human-in-the-loop methods.
<br>

• High-Volume Data Processing: Real-time streaming data ingestion, cleaning, delimiting, deduplication, and recalibration.
<br>

• Autoscaling for ML Workloads: Handling of large-scale batch jobs, job queues, and unexpected demand surges by scaling compute resources.
<br>

• Hypothesis A/B Testing: Design of experiments and execution. Make decisions about variable selection, control groups, and segmentation.
<br>
<br>
<div style="text-align: center;"> . . . </div>
<br>
&nbsp;&nbsp;  I believe in everyone's innate potential as they each have a vast cosmos within them.  Working with exceptional people and solving hard problems gives me the greatest joy.
<br><br><br>
</div>

.content-description {
    color: #4d455d;
    text-align: left;
    line-height: 1.5;
    padding-left: 100px;
    padding-right: 100px;
  }

@media (max-width: 767px) {
    .content-description {
      padding-left: 10px;
      padding-right: 10px;
    }

/* Adjust video container for mobile */
    .video-container {
      padding: 10px;
      width: auto; /* Adjust width to be responsive */
    }

/* Add any other mobile-specific style adjustments here */
  }

</style>