Dev Ambani Email & Phone Number

Evanston, Illinois, United States

Evanston, Illinois, United States

- Fine-tuned an open-source language model for legal contract drafting, utilizing topic modeling and text generation to boost performance by 25%.- Developed a Named Entity Recognition (NER) benchmark system on Amazon EC2, leveraging semantic analysis and GloVe embeddings, which increased evaluation accuracy by 30%.- Created APIs on AWS Lambda to support extensive model testing, achieving a 95% reliability rate for tasks like text summarization and information retrieval.- Optimized AWS data storage and retrieval processes with S3 and RDS, reducing retrieval time by 20% for model training and deployment.

Evanston, Illinois, United States

Evanston, Illinois, United States

Evanston, Illinois, United States

Evanston, Illinois, United States

- Co-Founder and CEO of Local, an app dedicated to connecting communities with local businesses through AI-driven solutions.- Leading a team of 30, alongside co-founder Saloni Patel, to innovate in urban exploration and local business engagement.- Secured over $10,000 in funding to support growth and expansion.- Serving over 300 local businesses with tools to boost visibility, customer engagement, and retention.- Nominated for the Chicago Innovation Awards, recognizing our impact in the local business space.- Part of The Garage, Northwestern University’s official incubator, fueling our mission with mentorship and resources.

Evanston, Illinois, United States

Mentored by Dr. Plamen Petrov, CEO of Healee· Creating an interactive dermatologist interface to enhance patient-doctor consultation practices.· Enhanced chatbot comprehension of patient issues by 40% using BERT, GPT, and Hugging Face Transformers for query analysis.· Utilized Inception V3 and few-shot learning to optimize ML-based image recognition, yielding 30% increase in diagnostic accuracy.· Crafted SpaCy-based NLP algorithms for sentiment analysis and topic modeling on patient feedback, yielding a 64% satisfaction rate.· Incorporated Bayesian Optimization to fine-tune hyperparameters, refining the model and achieving a notable 86% accuracy.

- Designed SentinAI to provide culturally sensitive, on-demand mental health support for underserved South African communities, addressing high suicide rates.- Developed an AI chatbot with 90% target accuracy for real-time suicide risk detection.- Built a suicide detection model using Python, TensorFlow, and Hugging Face Transformers within a RAG framework, leveraging Elasticsearch for precise responses.- Achieved 92% detection accuracy, contributing to a 35% increase in early interventions and earning endorsements from national mental health organizations.

Mumbai, Maharashtra, India

Mentored by Dr. Sunil Karamchandani, Senior IEEE Member· Developed a novel Python traffic system, using Snowflake and Azure for 40% quicker cross-validation of dynamic image processing.· Applied OpenCV & YOLO v5 model that communicated information to on-base Raspberry Pi leveraging Google Cloud GPUs & TPUs.· Improved real-time classification algorithm using CNTK and Keras, resulting in a 15% decrease in traffic congestion.· Initiated deployment of CNN model in cloud-edge infrastructure, facilitating a 25% increase in responsiveness during peak traffic.

Sunnyvale, California, United States

Mentored by Co-founder Samir Gandhi· Designed deep learning models through CNNs, fortifying customer behavior prediction and raising forecasting accuracy to 93%.· Implemented reinforcement learning in PyTorch for dynamic pricing, optimizing revenue growth by 47% in competitive markets.· Streamlined ML workflows on AWS, reducing data processing time by 54%, aiding in data mining and improving fraud detection.· Engineered recommendation systems utilizing collaborative filtering, escalating customer retention and repeat purchases by 25%.· Leveraged Docker for containerization and GitHub Actions to refine CI/CD pipelines, ensuring reliable and efficient deployments.

Mumbai, Maharashtra, India

Mentored by Wesley Weel, Director of AI Solutions for Automation· Led development of the Apriori Algorithm in R, increasing network reliability by forecasting system errors with 80% accuracy.· Automated data preprocessing for better accuracy using TensorFlow and Pandas, reducing manual data cleaning efforts by 42%.· Devised real-time financial anomaly detection using Scikit-Learn and XGBoost, detecting fraudulent transactions with 95% accuracy. · Employed ARIMA and LSTM for stock price forecasting, exceeding baseline models by 15%, enabling proactive investment strategies.· Elevated decision making by 42% through Power BI, facilitated via AWS Glue, leveraging CRM data within SAP ERP systems.

Master'S Degree, Artificial Intelligence, 4.0 / 4.0

Minor Degree, Engineering/Industrial Management

Bachelor Of Technology - Btech, Electronics, 3.9 / 4.0

Activities and Societies: IETE Member: Engaged in telecom sessions, expanded industry connections. CSI Member: Participated in tech.

High School Diploma

Middle School Diploma