
1
1
United States
4+ years of experience
I'm an AI/ML Engineer with 3+ years of experience across the full model lifecycle — data engineering, feature design, model training, and cloud deployment. I specialize in LLMs, RAG architectures, and applied machine learning on large-scale datasets, with a strong grounding in data analytics (Python, SQL, A/B testing, KPI reporting). Recent highlights: • RAG chatbot on AWS Bedrock → 71% to 88% accuracy, 99.4% uptime, serving 200+ advisors • Automated data quality checks → 63% reduction in pipeline errors • Lead scoring model → 19% increase in qualified conversions • Co-authored BARTNet (IEEE CAI 2025) — deep learning for real-world transit forecasting Stack: Python · SQL · LangChain · LlamaIndex · AWS Bedrock · PyTorch · Scikit-learn · Tableau

PulseSignal is a next-generation Market Intelligence platform built for Go-To-Market (GTM) and Sales teams. The platform solves the problem of 'lagging indicators' by treating unstructured web data—specifically high-velocity hiring and news—as leading indicators of corporate strategy. Our autonomous pipeline executes in three stages: 1. Live Data Intercept: Powered by Bright Data SERP API, we perform targeted crawls of Google Search and career pages. 2. Data Structuring: Google Gemini Flash parses messy HTML into strict JSON, identifying niche skills and business priorities. 3. Executive Synthesis: We implemented a Hybrid Multi-Model Architecture, routing complex reasoning through DeepSeek V4 Flash (via AI/ML API) to generate persona-specific playbooks for Sales, Recruiting, and Investors. Key features include a dynamic 'Manager's Weekly Report' with an Account Prioritization Matrix, an Evidence Ledger for transparency, and a 'Magic, Moment' Gmail integration that takes an AI-drafted email and pre-loads it, directly into the user's browser, ready to send. PulseSignal doesn't just show data; it generates quantifiable ROI.
31 May 2026