AI & Data Strategy Engineer

DEEP PATEL

Hasbrouck Heights, NJ 551-***-**** ***************@*****.*** LinkedIn GitHub

AI & DATA STRATEGY AI & SOFTWARE ENGINEERING DATA SCIENCE & ANALYTICS AI Engineer combining deep technical experIse with strategic business acumen. Dedicated to building producIon-ready agenIc AI systems and deep learning pipelines that transform raw data into scalable soluIons for complex market challenges. EDUCATION

RUTGERS UNIVERSITY, The State University of New Jersey, New Brunswick, NJ Master of Business and Science (MBS), Analy@cs: Data Science & AI – GPA 3.7/4.0 May 2026 Relevant Coursework: AI Engineering, AI: Concept to Market, AnalyIcs & Discovery InformaIcs, Advanced AnalyIcs PracIcum, Cloud and Big Data Systems, Business Intelligence, Science and Technology Management Capstone Bachelor of Science (BS), Computer Science – Minor in Economics and Business Administra@on – GPA 3.3/4.0 Jan 2025 Relevant Coursework: Deep Learning, Algorithms, Systems Programming, Internet Technology, Data Structures, Discrete Structures I & II, Data Management, Databases, SoZware Methodology, Econometrics, Financial Economics SKILLS

Programming & Data Engineering: Python, SQL, Java, JavaScript, C, Pandas, NumPy, PySpark, Hadoop AI/ML: PyTorch, TensorFlow, Scikit-learn, Hugging Face, LangGraph, Ollama, PydanIc, Deep Learning, AgenIc Systems SoRware Engineering: FastAPI, Flask, Django, Node.js, React, REST APIs, Cloud, Data Infrastructure & Deployment: AWS, Google Cloud Pla_orm (GCP), BigQuery, MongoDB, Docker, Git Automa@on & Data Intelligence: Selenium, Web Scraping, Workflow AutomaIon, Tableau, Looker PROFESSIONAL EXPERIENCE

Collabora@ve Solu@ons Grade Crossing Risk via Rutgers MBS Externship Exchange – Lead Extern Jan 2026 – Apr 2026

• Architected end-to-end machine learning predicIve pipelines uIlizing Python, XGBoost, and deep learning algorithms to analyze large-scale datasets and idenIfy high-risk railroad crossings.

• Built robust evaluaIon frameworks for heavily imbalanced safety data using Scikit-learn and TensorFlow, uIlizing AUCPR metrics to balance the costs of missed risks and unnecessary intervenIons.

• OpImized high-dimensional feature spaces via Principal Component Analysis (PCA), successfully reducing 36 demographic variables to 25 principal components while retaining 95% of data variance.

• Awarded the DisInguished Team Award out of 60+ teams for excepIonal technical execuIon and data integrity. DAFRE Climate Dashboard via Rutgers MBS Externship Exchange – Lead Extern Jan 2025 – Apr 2025

• Led the end-to-end engineering of an interacIve climate-health data dashboard, translaIng complex datasets into accessible visualizaIons while consistently hiing aggressive project milestones.

• Leveraged Python and R for rigorous data cleaning and staIsIcal analysis on large-scale environmental datasets, generaIng acIonable data science insights for policymakers. PROJECTS

Mul@-Agent Investment Research PlaZorm Jun 2026

• Engineered a producIon-grade agenIc AI pla_orm uIlizing Python and LangGraph to simulate an insItuIonal investment commijee, orchestraIng five specialized LLM agents to generate data-driven equity research.

• Reduced analysis latency by over 60% and API token consumpIon by ~40% by decoupling determinisIc data acquisiIon from LLM tool-calling workflows and implemenIng parallelized data pipelines using Selenium and external APIs, achieving sub-3-second Ime-to-first-token performance.

• Developed a real-Ime observability interface in React and FastAPI that streams agent decisions, execuIon traces, and research results, providing full transparency into mulI-agent reasoning workflows.

• Architected an autonomous peer-review framework where a Risk Agent criIques peer outputs through bounded revision cycles, while enforcing structured outputs via PydanIc schema validaIon to improve reasoning consistency, reduce quanItaIve errors, and ensure data-grounded analysis. AudioSketch – AI Mood-Based Content Genera@on PlaZorm Apr 2025 – May 2025

• Engineered an end-to-end machine learning pipeline in Python, seamlessly integraIng audio preprocessing, TensorFlow inference, and the Imagen3 API via Google Cloud Pla_orm (GCP) and Vertex AI into a user-facing Flask web applicaIon.

• Trained and opImized deep learning models on labeled audio datasets, achieving 92% classificaIon accuracy through rigorous feature engineering and model tuning. Mangrove – Co-Founder Feb 2025 – Apr 2025

• Co-founded an AI-powered producIvity startup and built a Chrome extension/Electron.js pla_orm to intelligently analyze content and organize browser tabs.

• Secured 2nd place and a $1,000 award at the Rutgers Scarlet Pitch entrepreneurship compeIIon against 50+ teams by demonstraIng a highly funcIonal, applied AI prototype.

Contact this candidate