Portfolio

Mohamed Ansar

Associate Solutions Engineer · Datasembly · Mississauga

System · Output · Signal

I help build and ship AI-powered applications with Python, ML, SQL, and SwiftUI.

Ex–Amazon (Prime Video)

Data

Production analytics

Snowflake, SQL, pipelines: outputs teams trust on a recurring cadence.

Product

Native iOS

SwiftUI and SwiftData: built for repeat use, not slide screenshots.

How I work

Clarity first

Stakeholder language up front; engineering that still holds after launch.

Profile

Technical builder at the intersection of data systems and delivery

Correct, explainable, operable, whether the interface is SQL, a dashboard, or TestFlight.

Most of my work sits where data meets decision-making: retail pricing-scale datasets, recurring pipelines, and analytics that teams run week after week, not one-off charts.

At Datasembly, that translates to deep SQL and Snowflake work, careful validation, stakeholder-ready reporting, and clean handoffs when scope moves.

I apply the same standard to software and applied AI: end-to-end ownership when the right answer is a product, not only a query. This portfolio is the overlap of what I ship for clients and what I build to push my craft.

Flagship

One product I took from screens in Figma-equivalent thinking through SwiftData models, TestFlight-style iteration, and App Store Connect submission.

SwiftUI · SwiftData · iOS 17+

Log sets and reps fast; see last session without digging.

Workout → exercises → sets (reps, weight). Local SwiftData graph, no account wall for core logging. I handled UX, implementation, icons/metadata, and submission, not a handoff to “someone else for the store.”

Submitted to App Store

What this demonstrates

  • Ownership from problem framing to App Store binary
  • SwiftData relationships that match real workout structure
  • UI density tuned for one-handed use between sets
  • Execution under Apple review and listing constraints

Shipped surface

  • Workout list, session editor, native date picker for new sessions
  • History visible in-flow so progression decisions stay quick
  • Local-first: data stays on device unless you add sync later

Stack

SwiftUI SwiftData iOS Xcode App Store Connect
Case study App Store

Screens

RepTrack empty workouts screen with Add Workout call to action
Workouts: empty state
RepTrack workouts dashboard with continue workout and weekly progress
Dashboard
RepTrack active workout session with exercise and set inputs
Active session
RepTrack workout session before exercises are added
Add exercise
RepTrack new workout screen with date picker
New workout: date

Archive

Selected technical work

SQL tools, pipelines, ML, retrieval, automation: samples of how I work end to end.

More projects

Earlier prototypes and depth. Expand for the full list.
SQL Browser App Learning Tool

SQL Playground: Snowflake style in your browser

A browser-based SQL practice environment using SQLite WASM with a lightweight Snowflake-style translation layer.

Product thinking, SQL fluency, and a useful learning interface.

Live demo
ML Pipeline Retries

TaskMaster: small ML pipeline with retries

A production-style mini pipeline with idempotent steps, retry logic, structured logs, and a live interface.

More than notebook ML: closer to system behavior.

Live demo
ML Retention Explainability

High-value customer predictor

A retention-focused model with validation, explainability, and basic MLOps hygiene.

Modeling with stakeholder usability in mind.

Live demo
Search LLM Local First

Grocery AI assistant: local semantic search

Offline product search and grounded Q&A using embeddings, FAISS, and a lightweight local model.

Retrieval, local AI, practical interfaces.

Live demo
NLP Transformers Sentiment

Amazon fine food sentiment

Baseline-to-transformer sentiment analysis with stronger NLP performance and deployment exploration.

Scraping Automation Python

Yahoo Finance news scraper

Headless browser scraper for JS-heavy news pages, normalized into a structured dataset.

Downstream NLP and research workflows.

Zero-shot LLM/NLI QA

Smart product categorization: zero-shot

Category validation with BART-MNLI and lightweight rules for low-label review.

EDA Python Visualization

Spotify music trends: quick EDA

Audio features and popularity over time, focused on clean data storytelling.

Sports Analytics Power BI Scoring Logic

Canadian Premier League: Dream XI (2019)

Role-aware KPI model for player selection with a Power BI dashboard for storytelling.

Capability matrix

Signal coverage

Warehouse SQL, orchestration, BI, and the product layers on top.

CH-01

Data systems

SQL, Snowflake, BigQuery, DuckDB

CH-02

Workflows

Airflow, ETL/ELT, recurring data delivery

CH-03

Analytics

Tableau, Power BI, pandas, business reporting

CH-04

Expansion

Python depth, AI workflows, search, automation, systems design

Timeline

Roles & impact

Solutions engineering and analytics operations: progressively more technical ownership from support through pre-sales solution design.

Associate Solutions Engineer · Datasembly

Jan 2026 – Present · Remote, Canada

SQL Snowflake Pre-sales Data Solutions
  • Design data solutions for retail pricing datasets, supporting client needs, pre-sales analysis, and custom data requests.
  • Use SQL and Snowflake to investigate complex data issues, validate outputs, and deliver reliable stakeholder-ready solutions.
  • Partner with internal teams to translate business questions into clear technical approaches and usable deliverables.

Tech Support · Datasembly

Aug 2024 – Dec 2025 · Remote, Canada

SQL Snowflake Reporting Workflow Support
  • Supported pricing analytics initiatives by building SQL-based reporting solutions and improving internal data workflows.
  • Resolved complex data issues across recurring client and internal requests using Snowflake and related tooling.
  • Contributed to dashboarding, reporting reliability, and workflow improvements across analytics operations.

Business Development Intern (Business Analysis) · Spongelii

Jan 2024 – Apr 2024 · Remote, Canada

Business Analysis SQL Snowflake Tableau
  • Business analysis tied to data workflows, reporting needs, and stakeholder requirements.
  • SQL and Snowflake-based work shaping dashboard and reporting outputs.
  • Documentation for requirements, analysis logic, and workflow structure.

Business Analyst II - Quality Auditing · Amazon Prime Video

Oct 2021 – Aug 2022 · Hybrid, India

Quality Auditing Process Improvement Reporting Operations
  • Quality auditing and process improvement for Prime Video content operations.
  • Reporting and analysis for workflow gaps, trends, and operational decisions.
  • SOP updates, audit consistency, cross-team standardization.

Business Analyst I - Digital Content · Amazon Prime Video

Aug 2019 – Sep 2021 · Hybrid, India

Digital Content Quality Control Stakeholder Coordination Operations
  • Quality control and operational coordination for the digital content catalog.
  • Stakeholder work on defects, metadata accuracy, and content readiness.
  • Process quality across recurring operations and issue resolution.

Contact

mohamedansarkms@gmail.com linkedin.com/kmsmohamedansar