Testimonials

What the people I have worked with said.

Alex Boches, CEO

Teal Omics

“Paritosh was instrumental in developing our AI-based multi-omics drug discovery platform. His goal was to use EHR datasets like UK Biobank and PPMI to identify key proteins and genes. He implemented techniques like GWAS and Mendelian Randomization. His work replicating research licensed from the Wyss-Coray lab at Stanford, and an in-house AI research analyst using fine-tuned LLMs with RAG, led to a 40% increase in potential drug targets and a 50% reduction in discovery time.”

AI-enabled drug discovery for age-related diseases

+40% drug targets · −50% discovery time

Ashley Blanchard, MD, and Max Topaz, PhD

NewYork-Presbyterian

“Paritosh developed risk detection models for pediatric care in the ER. His custom embedding models, using LSTM and SVM with neural networks, stratified children at high risk for abuse, neglect, and disease in real time. That led to a 30% improvement in early intervention rates and a 25% reduction in overlooked cases.”

AI in emergency pediatric care

30% better early intervention

Member of Technical Staff

Testudo Global

“Architected a production risk-assessment pipeline classifying 1M+ legal records and Temporal multi-agent enrichment across 8,000+ companies—cutting underwriting from weeks to hours.”

Lloyd’s-backed insurance underwriting

1M+ records · 95%+ accuracy

“Paritosh’s PubMed parsers and GraphRAG systems changed our approach to biotech investments. His methods improved investment accuracy by 35% and reduced analysis time by 60%.”

Biotech investment

+35% investment accuracy

Weber

FloraAI

“Paritosh’s work deploying and customizing Stable Diffusion models and LLMs was crucial to our launch. His full-stack contributions reduced our dependence on external API providers by 70%, a 40% cost reduction, and his testing and go-to-market work contributed to a 50% increase in user acquisition in the first quarter.”

AI-powered creative tools

70% less dependence on outside APIs

“Paritosh’s emergency detection system analyzed radio audio in real time for schools and airports. His ranking modules improved algorithm accuracy by 45%, a 30% faster response in critical situations, and a 20% reduction in false alarms.”

Real-time audio for emergency detection

45% more accurate · 30% faster response

Jake Cronin

Siro AI

“Paritosh’s work on audio AI, using MFCC and tonality analysis with transcription embeddings, changed how we identified sales strategy. The dashboard increased clients’ sales performance by an average of 25% and improved team efficiency by 30%.”

Sales performance from recorded calls

+25% sales performance

Wissam Akra

ToughLeaf

“Paritosh’s similarity search changed how we match clients and subcontractors. It increased matching accuracy by 40% and cut bidding time by 50%.”

Construction bidding

40% better matching