GTM & Marketing Analytics · AI Agent Builder · Production LLM Systems

Tasmia
Afroze

Turning commercial complexity into decisions and deploying AI that makes data answer back.

6+ years driving GTM, marketing, and revenue analytics across $1.3B+ portfolios and 35M+ users. Built and deployed two production AI agents on the Anthropic Claude API: a multi-touch GTM attribution engine and a natural-language-to-SQL analytics agent. MBA + MS in MIS, University of Pittsburgh Katz.

GTM & Marketing Analytics Multi-Touch Attribution Modeling AI Agent Development (Claude API) Forecasting & Revenue Analytics Snowflake · dbt · Python · Power BI Data-Driven Decision Enablement
View Projects Get in Touch
Tasmia Afroze
$1.3B+
Annual pricing & revenue portfolio
35M+
Customers analyzed for segmentation & churn insights
+15%
Forecast accuracy improvement delivered

About Me

Clarity from complexity,
decisions from data

📍 Pittsburgh, PA | Open to relocation

I'm a GTM and marketing analytics professional with 6+ years turning complex, fragmented commercial data into decisions that move the needle, from driving a 5% quarterly revenue lift on a $1.3B+ portfolio to building churn models across 35M+ subscribers, to deploying production AI systems on the Anthropic Claude API. My work spans B2B industrial (Carmeuse), healthcare (Baxter International), and telecom (Robi Axiata), which means I bring both analytical depth and cross-domain adaptability.

What sets me apart is that I build with AI, not just use it. I shipped two production AI agents: a multi-touch GTM attribution engine comparing Last Touch, Linear, Markov Chain, and Shapley Value models with a Claude API layer that auto-generates executive narratives and budget recommendations, and a natural-language-to-SQL agent that lets non-technical stakeholders query data warehouses in plain English. Both are live on GitHub. I hold an MBA + MS in MIS from the University of Pittsburgh (Katz), a CSPO, and a Lean Six Sigma Green Belt. Open to relocation.

Featured Work

Projects & Portfolio

AI / LLM Engineering

Natural Language → SQL Agent

Built a production AI agent that converts plain English business questions into optimized SQL queries using the Anthropic Claude API, executes them against a real sales database, and returns results with a plain-language explanation and automatic data visualization.

Key Insight: Combining LLM reasoning with structured data execution eliminates the technical barrier between business questions and database answers — no SQL knowledge required.
Claude API Live demo
Python Anthropic Claude API SQL SQLite Streamlit LLM Integration

Python · AI · GTM Analytics

Multi-Touch GTM Attribution Engine

Built a B2B GTM attribution engine comparing Last Touch, Linear, Markov Chain, and Shapley Value models across marketing touchpoints. A Claude API layer auto-generates executive narratives and budget recommendations directly from attribution outputs, turning model results into board-ready insights without analyst intervention.

Key Insight: Shapley Value attribution revealed that mid-funnel content touches were systematically under-credited by last-touch models, shifting recommended budget allocation toward nurture programs by approximately 30%.
4 attribution models AI-generated narratives Budget recommendations
Python Anthropic Claude API dbt Snowflake Markov Chain Shapley Value LLM Integration

Power BI · Marketing Analytics

Multi-Channel Pre-Lead Performance Dashboard

Built a monthly pre-lead dashboard consolidating email marketing, SEO, paid digital advertising, and website engagement data for two B2B clients. Enabled side-by-side competitor benchmarking, campaign-level CPL and cost-per-conversion tracking, and source-level website attribution to guide monthly spend allocation decisions.

Key Insight: Email visitors drove 3× more pages per session than social traffic, signaling higher intent — a finding that directly informed channel investment priorities.
4 channels tracked 2 clients benchmarked CPL · CTR · Conversion
Power BI DAX Email Analytics SEO Analysis Digital Advertising B2B Marketing

Rebuilt with anonymized data to protect client confidentiality.

EDA & Python

AI-Driven Pricing Simulation & Discount Optimization

Analyzed a $2.3M retail dataset to identify profit leakage drivers and built a decision tree-based simulation model to evaluate counterfactual pricing scenarios. Simulated discount optimization scenarios, estimated ~$182K profit uplift by capping excessive discounting and preserving margins on low-performing products.

Key Insight: Profitability is driven by nonlinear interactions between discounting, product mix, and volume, requiring targeted, data-driven pricing decisions.
estimated ~$182K profit uplift $2.3M dataset analyzed
Python pandas Decision Tree Simulation Modeling Google Colab

Forecasting & Time-Series ML

Revenue Forecasting Dashboard

Built an end-to-end revenue forecasting application using Facebook Prophet with seasonality tuning and scenario modeling. Layered an interactive Streamlit + Plotly dashboard on top — executive KPI cards, confidence interval charts, product-line drill-down, and what-if scenario comparison views.

Key Insight: Decomposing seasonality and trend separately lets business users run scenario comparisons (Base Case vs. Optimistic vs. Conservative) without touching the model — turning a data science output into a planning tool.
90-day horizon 3 forecast scenarios Live demo
Python Prophet Streamlit Plotly pandas Time-Series Forecasting

Machine Learning

Customer Churn Prediction Model

Designed and deployed a Python-based churn prediction pipeline using scikit-learn on 35M+ subscriber records. Implemented lifecycle segmentation and A/B tested loyalty interventions that reduced churn by 8% and prevented $200K in revenue loss.

Key Insight: Early-stage engagement signals were the strongest churn predictors, enabling proactive intervention before intent to leave solidified.
8% churn reduction $200K loss prevented
Python scikit-learn pandas A/B Testing Customer Retention Lifecycle Strategy

Healthcare Analytics

Patient Outcomes Dashboard for AE Enablement

Built dashboards tracking 100K+ patient profiles to surface adoption patterns and clinical outcomes. Trained 30 Account Executives on data-driven selling, reducing hospital admissions by 60% and cutting unnecessary inquiries by 18%.

Key Insight: AEs with access to outcome data closed consultations faster and with higher confidence, making data literacy a direct revenue lever.
60% lower admissions 30 AEs trained
Tableau SQL CRM Analytics

Client Consulting Project · Six Sigma

Project Lead – Six Sigma Process Improvement | ABARTA Coca-Cola Beverages

Led a client-facing Six Sigma process improvement initiative for ABARTA Coca-Cola Beverages, redesigning rework processes to reduce breakage damage and loss (BDL) over-expenditure. Drove root cause analysis and cross-functional process redesign, identifying inefficiencies in rework workflows as the primary driver of cost leakage. Delivered a structured improvement plan with ~$48K projected savings over six months, demonstrating measurable impact through process optimization rather than volume reduction.

Key Insight: Rework process inefficiencies were the root cause of BDL over-expenditure, addressable through structured process redesign rather than volume reduction.
~$48K projected savings 6-month timeline
Lean Six Sigma Process Improvement Root Cause Analysis Team Leadership Stakeholder Management

Client Consulting Project · Digital Marketing

Consultant (MBA Engagement) | HENY

Designed and executed a data-driven digital marketing strategy, increasing website traffic by 10% through targeted content and audience segmentation.

Key Insight: Targeted content strategy drove measurable traffic growth, demonstrating that audience-aligned messaging outperforms broad digital spend.
10% website traffic increase
Digital Marketing Content Strategy Audience Segmentation Technology Consulting

Client Consulting Project · Finance

BNY Mellon Fellowship | Birgo Realty

Developed a client-facing strategy for Birgo Realty to address financial literacy gaps among rental tenants, identifying opportunities to reduce risk exposure and improve tenant stability. Delivered recommendations to stakeholders, highlighting opportunities to reduce tenant default risk and improve financial engagement.

Key Insight: Financial literacy gaps represent both a social impact opportunity and a lever for reducing tenant default risk.
Jan 2022 – Apr 2022
Case Study Development Financial Literacy Client Consulting Strategy

Technical Toolkit

Skills & Tools

Tools and frameworks I use to drive pricing, growth, and customer analytics decisions.

🎯

Business Strategy & Pricing

Pricing Strategy Revenue Optimization Customer Segmentation Growth Strategy Experimentation (A/B Testing)
📊

Analytics & Data Foundations

SQL Python pandas Data Modeling A/B Testing
🤖

Machine Learning & AI

Predictive Modeling Customer Analytics Model Evaluation Feature Engineering Applied Machine Learning Churn Prediction
📈

Business Intelligence

Power BI DAX (Advanced) Tableau Snowflake SAP Tagetik EPM
🔗

CRM & GTM Tools

Salesforce CRM Sales Funnel Analytics Forecasting & Planning Sales Enablement
🌍

Domain Expertise

Industrial B2B Healthcare Analytics Telecom/Lifecycle Marketing Analytics Sales Strategy

Career History

Work Experience

Jul 2023 – Apr 2026

Marketing & Sales Data Analytics

Carmeuse — Pittsburgh, PA

  • Delivered a 5% quarterly revenue lift by identifying discount inconsistency (not product mix or volume) as the primary margin erosion driver through price-volume-mix decomposition, realigning pricing governance across the $1.3B+ portfolio
  • Improved forecast accuracy by 15% by building Python (Prophet) forecasting frameworks and scenario models, directly enabling faster executive resource-allocation decisions
  • Designed and deployed a production NL-to-SQL agent on the Anthropic Claude API (Streamlit, Snowflake, human-in-the-loop validation) enabling non-technical stakeholders to query the data warehouse in plain English — eliminated 10+ hours/week of manual ad-hoc reporting
  • Built dbt semantic models and automated reconciliation across Salesforce, SAP, and Snowflake, creating a single source of truth for Sales, Finance, and Operations leadership
  • Delivered 7% operational efficiency gains through scalable SQL pipelines, Power BI dashboards (advanced DAX), and data governance standards across the commercial data layer

May 2022 – Dec 2022

HealthTech Analytics Intern

Baxter International

  • Built patient outcomes dashboards tracking 100K+ patient profiles, enabling data-driven clinical decision support
  • Trained 30 Account Executives on analytical selling, contributing to 60% lower hospital admissions in tracked cohorts
  • Reduced unnecessary support inquiries by 18% through proactive analytics-driven escalation workflows

Jul 2018 – Jun 2021

Product Marketing Manager (Customer Lifecycle Management)

Robi Axiata — Dhaka, Bangladesh

  • Led lifecycle marketing and retention strategy for 35M subscribers, leveraging churn modeling and segmentation to drive targeted engagement and revenue protection
  • Reduced churn by 8% and protected $200K in revenue by transforming stakeholder hypotheses into structured A/B testing frameworks
  • Designed loyalty program architecture that drove a 12% increase in customer engagement metrics

Academic Background

Education

2021 – 2023

MBA, Marketing & MS, Management Information Systems

University of Pittsburgh — Joseph M. Katz Graduate School of Business

Focused on analytics, pricing strategy and data-driven decision making across marketing and operations.

President, Business Technology Club: led initiatives bridging business and analytics, driving engagement in data-driven decision making through industry sessions and case competitions
Co-President, Finance Club: organized industry sessions and peer learning initiatives
Student Body Representative, Super Analytics Challenge 2023: elected to serve as voice of students in designing the third annual Super Analytics Challenge; curated the Challenge dataset and wrote the case on Mental Health Challenge while providing feedback across all logistical specifications

Honors & Awards

Recognition

Dean's Outstanding Full-time MBA Student Service Award
Winner, Katz Invitational Case Competition
Runners-up, Super Analytics Challenge 2022

2014 – 2017

BSS, Economics

University of Dhaka, Bangladesh

Built strong foundation in econometrics, statistics, and quantitative analysis.

Associate Member, Economics Study Center (ESC)
Winner, EconProdigy 2.0

Certifications

Professional Credentials

Industry-recognized certifications

Certified Scrum Product Owner (CSPO)
Lean Six Sigma Green Belt
Google Data Analytics Professional Certificate

Let's Connect

Get in Touch

I'm a GTM and marketing analytics professional who builds production AI systems, not just reports. I've deployed two AI agents on the Anthropic Claude API, owned analytics across a $1.3B commercial portfolio, and driven measurable outcomes across B2B industrial, healthcare, and telecom. Currently exploring senior analytics, GTM strategy, and AI-native analytics roles where data and AI actively shape business decisions.

✉️ Email Me 💼 LinkedIn 🐙 GitHub

Always happy to connect about analytics, pricing strategy or building data-driven products.

📍 Pittsburgh, PA  ·  Open to relocate