Chandra Prakash

Founder

OneSpotAI

Dec 2025 - Current (9 months)

Building a full-stack AI consulting firm delivering end-to-end AI solutions and taking complete ownership of engineering for startups and enterprises across domains. - Leading development of custom AI agents and end-to-end AI systems, leveraging LLMs, reinforcement learning, and domain-specific modeling for real-world applications. - Delivering production-ready AI products at high velocity through lean, AI-first engineering practices, enabling significantly faster iteration compared to traditional teams. - Owning the full lifecycle from architecture and development to deployment and scaling, including full-stack applications, APIs, and infrastructure. - Partnering closely with clients across healthcare, finance, and other domains to solve

Senior Machine Learning Engineer

Stealth Startup

Sep 2025 - Current (1 year)

Leading the development of AI agent systems and Predictive Models for Finance use cases, delivering production-ready solutions for decision-making and automation. -Built and optimized multi-agent architectures for financial workflows, focusing on scalability, robustness, and real-world deployment. -Improved agent accuracy and reliability through fine-tuning and alignment techniques, along with orchestration-level enhancements across agent interactions. -Developed in-house machine learning methods for predictive pricing tailored to complex financial datasets. -Designed and deployed a Bayesian probabilistic model for price prediction, enabling reliable forecasts despite noisy and incomplete financial data.

Senior Machine Learning Engineer

TheAgentic

Aug 2025 - Jan 2026 (6 months)

Led the development and deployment of production-ready AI systems for clients across healthcare, finance, and industrial domains, delivering end-to-end machine learning solutions across multiple high-impact projects. - Built an end-to-end agentic pipeline for computational protein antibody generation against target antigens, leveraging protein structures, amino acid sequences, and signals such as immunogenicity and binding affinity for candidate design and optimization. - Architected a Primary Source Verification (PSV) platform for US medical practitioners using browser automation and multi-channel communication agents across phone, SMS, and fax, supported by robust orchestration logic. - Designed an AI-driven document intelligence pipelin

Research Intern

IRT group AI Institute of South Carolina

Oct 2024 - Mar 2025 (6 months)

Conducted research on large language model alignment and fine-tuning, focusing on improving model reliability, adaptability, and knowledge retention. - Explored active learning strategies to identify high-value data points for training, improving data efficiency and reducing labeling overhead. - Investigated continual learning approaches for transformer models to enable incremental knowledge updates without full retraining. - Developed techniques to incorporate new knowledge into LLMs without increasing model parameters, while mitigating catastrophic forgetting and preserving previously learned capabilities.

Founding Machine Learning Engineer

ZynixAI

May 2024 - Jul 2025 (1 year 3 months)

Joined as the founding ML Engineer and built the company’s machine learning infrastructure from scratch. - Developed a Transformer-based model on raw EMR data to predict readmission, mortality, and clinical risks with over 90% accuracy. - Built Voice AI agents for scheduling and follow-ups, integrated with a custom MCP Server to maintain patient context and update hospital/ACO systems. - Designed RAG pipelines on EMR and claims data, achieving over 75% accuracy in HCC suspect coding to support value-based care. - Automated SOAP note generation from audio transcripts, improving information recall by 40%. - Fine-tuned LLMs on proprietary healthcare data to enhance clinical NLP performance. - Built a text-to-SQL engine with MCP Server integra

Research Intern

UiT The Arctic University of Norway

Apr 2024 - Sep 2024 (6 months)

Conducted research on large language model quantization, focusing on understanding the impact of reduced precision on model behavior and performance. - Analyzed the fundamental differences between quantized and full-precision models, studying their effects on accuracy, stability, and inference efficiency. - Evaluated existing quantization techniques and identified limitations affecting reliability in real-world deployments. - Explored methods to improve the accuracy and robustness of quantized models while maintaining computational efficiency and scalability.

Research Intern

Flamelai

Feb 2024 - Apr 2024 (3 months)

Conducted research and development on advanced generative modeling techniques, leveraging state-of-the-art latent diffusion models and diffusion transformers to address complex image synthesis challenges. - Experimented with latent diffusion and diffusion transformer architectures to improve image quality, controllability, and generation efficiency. - Explored conditioning mechanisms and training strategies to enable more precise and context-aware image synthesis. - Evaluated model performance across diverse datasets and tasks, focusing on improving robustness, scalability, and visual fidelity.

Research Intern

Indian Institute of Science IISc

May 2023 - Jul 2023 (3 months)

Worked on generative approaches for solving highly non-linear partial differential equations, exploring the use of differential equation GANs and diffusion-based methods. - Developed Physics-Informed Neural Network and Generative Adversarial Network hybrid models to learn solutions to complex non-linear PDEs. - Investigated diffusion model approaches for solving partial differential equations, focusing on improving stability and solution quality for challenging regimes. - Translated research ideas into code, trained and evaluated models, and contributed to experimentation and benchmarking across multiple PDE settings.

Research Intern and Mentor

Indian Institute of Technology Indian School of Mines Dhanbad

Nov 2022 - Apr 2024 (1 year 6 months)

Worked on applying computer vision and generative modeling techniques to domain-specific challenges in geophysics and medical imaging, collaborating closely with faculty on research problem statements. - Assisted professors in translating domain problems into machine learning solutions, developing and evaluating models tailored to geophysical and medical imaging data. - Built and experimented with computer vision and generative models to improve analysis, reconstruction, and interpretation of complex visual data. - Served as a founding member of the AI Club, mentoring students and guiding them to get involved in machine learning research and projects.

About

Detail

Timeline

Résumé