About me

Hello! I’m Shamika Likhite, a Software Engineer (Machine Learning) at Speechace, LLC.

At Speechace, I focus on developing and enhancing machine learning models for our Writing Assessment API, with work spanning both ML engineering and data engineering. My contributions aim to organize, optimize, and advance the speech and writing assessment capabilities of Speechace. Specifically, I design machine learning and natural language processing (NLP) solutions to address real-world challenges in text analysis and audio processing. These applications are widely deployed but often face hurdles related to data quality and model robustness. My goal is to build reliable, efficient algorithms that perform well in resource-constrained environments while maintaining high accuracy.

Beyond writing assessment, I am deeply interested in multimodal learning for NLP—developing models that integrate speech, text, and video to improve robustness, interpretability, and real-world applicability of natural language systems.

My broader research interests lie in long-context multimodal LLM reasoning—developing methods that allow large language models to effectively integrate and reason across multiple modalities (speech, text, video) over extended contexts. I am particularly interested in approaches that enhance robustness, interpretability, and efficiency in these systems.

Before joining Speechace, I worked as a Research Assistant at Northwestern University’s Electrical and Computer Engineering Department. There, I developed a multimodal model for speech disfluency detection at the Ideas Lab, focusing on robustness to missing modalities. I also contributed to video super-resolution research at the Image and Video Processing Lab (IVPL).

I hold a Master’s degree in Computer Engineering from Northwestern University.

📧 You can reach me at shamikalikhite2024 [at] u [dot] northwestern [dot] edu.

Here is my current work at Speechace that focuses on Writing Assessment

To learn more, please take a look at my CV

🗞️ News

• Sept 2025: I am currently looking for a new role in the industry.
• Sept 2024: Started working at Speechace.
• Aug 2024: Released a new Pypi package Video Audio recorder for Multimodal dataset creation on edge devices.
• June 2024: Our paper Missingness-resilient video-enhanced multimodal disfluency detection is accepted at Interspeech 2024.

🎓 Education

• M.S. in Computer Engineering, Northwestern University (Sep 2022 – Dec 2023)
• B.S. in Electrical Engineering, Visvesvaraya National Institute of Technology (July 2018 – May 2022)

👨🏻‍💻 Work Experience

• Software Engineer, Machine Learning, Speechace LLC (Sept 2024 – Present)
• Research Engineer, Northwestern University (Jun 2023 – Sept 2024)
• Machine Learning Engineer Intern, Merkle Inc (May 2021 – Aug 2021)