Available to hire
Hi, I’m Siddharth Kushawaha, a machine learning engineer specializing in turning cutting-edge research into scalable, production-ready AI products. I have 3+ years of experience building end-to-end AI systems with a focus on NLP, computer vision, and large language models. I’m proficient in Python, TensorFlow, and PyTorch, with hands-on experience in data engineering, MLOps, and model evaluation. I thrive in fast-paced environments and enjoy deploying high-performance models on AWS, Azure, and GCP to deliver meaningful product impact.
Outside of coding, I enjoy collaborating across teams to design robust data pipelines, optimize model performance, and mentor others on best practices in ML operations and responsible AI deployment.
Skills
Language
Afar
Advanced
Work Experience
Data Analyst at Community Dreams Foundation
September 1, 2025 - PresentOptimized complex SQL queries by indexing high-usage columns, rewriting subqueries, and leveraging partitioning strategies, resulting in a 60% reduction in query time and faster access to campaign performance metrics. Designed an agentic AI workflow (LangGraph) that queries Snowflake, Hudi, and AWS RDS, integrating a chatbot to interpret user queries and retrieve business insights, enhancing internal stakeholder self-service and reducing ad-hoc reporting requests by 50%.
Data Scientist at Factspan Analytics
June 1, 2022 - June 1, 2023Designed an architecture to handle concurrent LLM inference requests using Kubernetes and vLLM, incorporating autoscaling and GPU optimization, and integrated benchmarking to monitor latency, throughput, and token generation time. Engineered a high-performance sorting solution by fine-tuning Co-DETR for visual object detection and segmentation, improving recycling efficiency by 15% and reducing misclassification errors by 22% compared to legacy vision-based systems. Optimized customer-facing virtual assistants by implementing Quantized Low-Rank Adaptation (QLoRA) on LLaMA-3.1-7B, reducing inference latency using vLLM by 30% while maintaining high response quality. Strengthened security and compliance across RAG pipelines by integrating guardrails, AWS Secrets Manager for credential management and Okta for role-based access, reducing unauthorized access risks.
Junior Data Scientist at Factspan Analytics
January 1, 2021 - June 1, 2022Designed and implemented A/B tests for sales campaigns, providing consultation to Product Marketing, resulting in a 15% increase in conversion rates through data visualization using Tableau dashboards. Automated client segmentation using clustering models retrained on continually updated client records, with MLflow versioning and rollback ensuring high-quality marketing insights and minimizing manual intervention through a robust, monitored model registry. Enhanced an alerting system by incorporating Monte Carlo simulations and non-parametric statistical tests to identify anomalies in irregular patterns in customer behavior, improving robustness and resulting in a 74% improvement in detecting data quality issues. Engineered prompt optimization strategies and chain-of-thought techniques for generative models, improving output quality by 45% and reducing hallucination rates by 30% in production environments.
Data Analyst at Factspan Analytics
January 1, 2020 - December 1, 2021Led a data migration project to transfer campaign data across Snowflake servers, automating SQL-based workflows with Python, reducing execution time by 70% and improving pipeline reliability. Identified a critical bottleneck and implemented a high-performance SQL module that improved transformation speed by 37%. Built Tableau and Looker dashboards tracking CTR, ROAS, and lifetime value for 147+ marketing channels, enabling data-driven budget shifts that improved overall campaign ROI by 22%.
Data Analyst at Community Dreams Foundation
September 1, 2015 - PresentOptimized complex SQL queries by indexing high-usage columns, rewriting subqueries, and leveraging partitioning strategies; reduced query time by 60% and enabled faster access to campaign performance metrics. Designed an agentic AI workflow using LangGraph to query Snowflake, Hudi, and AWS RDS, integrating a chatbot to interpret user queries and retrieve business insights, increasing stakeholder self-service and reducing ad-hoc reporting requests by 50%.
Education
Master of Science in Data Science (Concentration in Business Analytics) at Stevens Institute of Technology
January 11, 2030 - December 1, 2025Master of Science in Data Science at Stevens Institute of Technology
January 11, 2030 - December 26, 2025Qualifications
Industry Experience
Software & Internet, Professional Services, Education, Media & Entertainment, Other
Skills
Hire a Data Scientist
We have the best data scientist experts on Twine. Hire a data scientist in Jersey City today.