Adam Dalal

Email: adalal80@gmail.com

LinkedIn: abdalal

Website: adalal80.github.io

Location: New York, NY

Phone: (832) 646-3179

Summary

AI Product Manager with a background in chemical engineering and economics, specializing in diagnosing where complex workflows break and building the right AI fix for that specific gap. I own AI/ML portfolios end-to-end, from problem framing and roadmap through governance and adoption, and still code the models I ship. My instinct is systems-first: find the constraint, isolate the variable, build for that. That approach has worked in financial services and I am looking for a senior AI role where the problems are hard and the outcomes are real, regardless of domain.

Technical Skills

Languages: Python, SQL
Platforms: Azure, Databricks, OpenAI API, Lovable
AI/ML: Prompt engineering, LLM development, model evaluation, GenAI pipelines, pandas

Experience

JP Morgan Chase — New York, NY

AI Product Manager | Sep 2019 – Present Control Management, Compliance, and Operational Risk — Internal Tooling/Platform

Owned a portfolio of 18 AI/ML decisioning products serving ~4,000 Control Managers responsible for operational and regulatory risk, replacing manual compliance workflows with automated intelligence.
Improved accuracy across high-criticality models by 5–15% and grew overall adoption by 30% by running feedback sessions with Control Managers to identify where model outputs created confusion, using those findings to de-prioritize retraining on low-impact models and redirect capacity toward higher-risk use cases, increasing trust and risk coverage where it mattered most.
Diagnosed a 45% QC rejection rate in the risk event loss platform as a submission standards problem rather than a data problem, Python-coded a validated GenAI prompt to draft and gap-check descriptions, and secured integration by influencing the application PM and feature team without owning their roadmap, reducing rejections to 15% and shortening approval cycles from 5 to 3 days.
Reduced regulatory review burden across 175,000 unstructured descriptions by designing LLM-based question and answering models sequenced by regulatory risk across 7 fields, improving compliance pass rates from ~96,500 to ~140,000 (45% increase) and cutting per-description review time from 3 minutes to 1 minute under a fixed regulatory deadline.
Partnered with Financial Crimes Monitoring analysts who were relying on regex rules that kept missing sensitive data, diagnosed the false negative problem as a model limitations issue, delivered a shared ML-based PII detection platform that improved detection accuracy by 40% and enabled autonomous scanning of ~2 million transactions.
Managed competing demand from 6 internal teams seeking access to the PII detection platform, prioritizing a focused credit card detection model for Financial Crimes Monitoring while pushing back on customization and fine-tuning requests that would have fragmented the model and pulled engineering capacity away from higher-impact roadmap commitments.
Served as the org’s de-facto AI SME across 18 products, setting 80% minimum accuracy thresholds and holding models from deployment until retraining met the bar, briefing the governance forum across Compliance, Legal, and Risk, and presenting deployment decisions to Control Manager leads.

Analytics Product Owner

Apr 2017 – Sep 2019

Owned product strategy and roadmap for firm-wide compliance and analytics products, translating regulatory mandates into automated solutions, prototyping features in Python, and building team capability spanning due date compliance, data quality monitoring, and analyst upskilling.
Spearheaded a Python-based due date engine automating compliance scheduling for 30,000 manually tracked due dates, reducing late controls from 10,000 to 2,500, then added an explainability layer after Control Managers said they trusted the outputs but needed to understand the reasoning behind them to fully adopt it.
Architected a compliance hygiene dashboard tracking 20 regulatory metrics for Control Managers, prototyped logic in Python against governance standards, reducing average exception age from 20 to 8 days and escalating aging exceptions to senior management to drive accountability where manual tracking had failed.
Mentored 4 analysts transitioning from Excel to Python, running bi-monthly sessions connecting JPMC learning coursework to hands-on data analysis with pandas, accelerating their ability to perform independent analysis without engineering support.

Senstay — Los Angeles, CA

Revenue Manager | Aug 2015 – Jun 2016 Property Management Company in Short-Term Rental Market

Owned product strategy for automated pricing and booking decisioning, building models that dynamically priced units on Airbnb and HomeAway and determined booking acceptance, increasing client revenue by ~20% and reducing operational costs by ~25% through workflow automation.
Led A/B testing and product experimentation to measure impact of professional photography and staging on engagement and revenue, analyzing results across identical units to inform product investment decisions and client recommendations.
Managed stakeholder relationships with property owner clients, conducting monthly performance reviews to present revenue reports, gather feedback on pricing strategies, and iterate on models based on client needs and market conditions.
Developed 5 market-entry analyses for a short-term rental platform, combining hotel and Airbnb data to evaluate city expansion viability and presented recommendations to the C-suite.

Lucky Vitamin — Los Angeles, CA

Senior Data Analyst | Feb 2015 – Aug 2015 Global Online Retailer for Health, Natural, and Organic Products

Designed and executed dynamic pricing experiments and category performance analysis to inform merchandising decisions.

Education

General Assembly — Data Science Certification New York, NY
Texas A&M University — MS Economics College Station, TX
Iowa State University — BS Chemical Engineering and Economics Ames, IA