Siddhant Gupta is an accomplished AI Engineer and Software Developer with expertise in large language models (LLMs), retrieval-augmented generation (RAG), software engineering, and backend development. With a strong background in AI research and full-stack development, Siddhant has successfully delivered scalable AI solutions, system optimizations, and high-impact software applications across various domains.

Currently, Siddhant is an AI Engineer at the Center for Applied AI, where he fine-tunes LLaMA 3.2 and Mixtral models using Supervised Fine-Tuning (SFT) and PEFT (LoRA, QLoRA) to optimize domain-specific AI applications. His work includes developing LLM-powered browser-based AI tools, leveraging OpenAI, Claude, and Meta AI models to create web-accessible, real-time AI solutions. Additionally, he has built scalable RAG pipelines using vector databases (Pinecone, ChromaDB) to reduce query latency by 40% while processing over 1M+ documents.

Beyond AI model development, Siddhant has demonstrated full-stack and backend expertise by engineering production-grade AI systems. He developed a mortgage-lending platform deployed on Azure, automating borrower eligibility, program matching, and real-time rate calculations, directly supporting a startup’s successful funding. He also built All Things Documents (ATD), a multimodal document intelligence platform leveraging Meta’s LLaMA 3.2 Vision model, which was adopted company-wide for automated insights and AI-driven document processing.

Siddhant’s experience extends to cybersecurity-focused AI research. As a Graduate Research Assistant at the Laboratory for Physical Sciences, he developed an AI-driven malware detection system, enhancing threat detection accuracy by 30% through optimized YARA rule scanning and automated data pipelines. His research led to the deployment of a real-time classifier with 95% accuracy, contributing to cutting-edge cybersecurity defenses.

Siddhant is passionate about bridging AI research with real-world applications, designing scalable browser-based AI tools, and optimizing LLM-powered workflows for enterprise and consumer AI solutions.

Ashutosh Latwala is a highly motivated AI Engineer and Associate Software Engineer with 4+ years of experience in data science, AI/ML, software/web development, and databases. Skilled in AI/ML, Python, AWS, and data analytics, specializing in creating robust, data-driven applications. Experienced in implementing AI models, AI applications, optimizing data pipelines, testing and delivering end-to-end solutions that drive efficiency and impact. Known for delivering scalable solutions and seamlessly integrating AI and software engineering expertise into impactful projects.

Presently, Ashutosh serves as an AI Engineer at the Center for Applied AI – UMBC Training Centers, where he develops and deploys AI-driven solutions like the Financial Data Analyzer and language models. He leverages AI assistants, APIs, machine learning techniques, and fine-tuning to build specialized models. Additionally, he optimizes AI systems by monitoring performance, conducting data-driven analysis, and applying advanced methodologies for continuous improvement. His work also involves enhancing AI research, integrating innovative solutions, and fostering collaboration through clear documentation and cross-functional teamwork.

Ashutosh also served as a Graduate Teaching Assistant at the University of Maryland, Baltimore County, where he supported students in Computer Science, Computer Engineering, and Information Systems departments. He enhanced the learning experience by providing grading, personalized feedback, proctoring exams, and conducting demo sessions, ensuring high-quality education and student engagement. 

Previously, Ashutosh worked as an Associate Software Engineer at Duma Infoservices Pvt. Ltd., where he led the development of web-based solutions, managed project documentation, and ensured timely delivery. He contributed to optimizing database management systems, improving data efficiency, and enhancing user experiences through intelligent recommendation systems. His work involved collaborating with cross-functional teams, refining processes, and implementing innovative solutions to enhance overall software & system performance.

Devon Slonaker is a dynamic and highly skilled professional with expertise in artificial intelligence, software engineering, and system administration. With a strong foundation in AI, including the use of embedding models, neural networks, and advanced machine learning techniques, Devon has excelled in research and product development across diverse AI and IT fields. His experience spans AI model development, system administration, and academic mentorship.

Currently, Devon serves as an AI Engineer at the Center for Applied AI, where he conducts research on AI techniques such as vector databases, language models, and AI applications in low-powered environments. He utilizes tools and techniques such as retrieval-augmented generation (RAG) and fine-tuning to create models specialized in specific tasks. He utilizes frameworks like PyTorch to develop neural networks and machine learning solutions, driving innovation in AI product engineering.

Devon has also demonstrated his skills as a Graduate Assistant at the University of Maryland – Baltimore County, where he assisted students with computer security and malware analysis coursework while developing teaching materials and ensuring effective student support. His work as a Linux System Administrator also involved managing critical infrastructure for UMBC’s Computer Science and Electrical Engineering department, leading a server migration from a RHEL to Ubuntu environment.

Devon’s technical expertise is further showcased through his hands-on project experience, which includes building a movie recommendation system, developing an encrypted peer-to-peer file system, and creating an Android app for in-game decision-making.