Skip to content

Applied AI & ML Engineer · Canada

Building AI products and ML systems with measurable impact.

My experience spans source-grounded LLM workflows, predictive models, evaluation systems, data and feature pipelines, APIs, and cloud delivery, with measured improvements in quality, efficiency, review effort, and decision support.

Explore experience Projects Contact
  • LLM systems, RAG & evaluation
  • Predictive ML & human review
  • Measured quality & efficiency gains

01 · Experience

AI and ML engineering grounded in professional delivery.

Applied AI evaluation, predictive ML, software and data delivery, financial technology, and public-sector research provide the professional foundation for the engineering work.

Current role

Data & Applied AI Analyst

BC Rapid Transit Company · Dec 2024 to present · Burnaby, BC

Build applied AI evaluation and human-review controls, predictive ML, and reusable data and feature foundations for enterprise decision support.

  • AI qualityBuilt a 150-case evaluation harness using RAGAS, semantic similarity, LLM-as-judge rubrics, trace capture, and latency and cost checks.
  • ML impactBuilt an uncertainty-aware recovery-duration model that improved planning accuracy by 21% through calibration and interval-coverage checks.
  • TrustReduced unsupported AI-generated summary claims by 60% while keeping human review and release boundaries explicit.

2023 · 2024

Brain Station 23

Data Scientist, Applied AI

Financial analytics, predictive ML prototypes, document intelligence, production-service contributions, data pipelines, and cloud integrations.

21%
analysis efficiency
17%
data consistency
26%
query performance

2023

UBC MDS × Statistics Canada

Capstone Data Scientist

Interpretable clustering, data profiling, outlier analysis, comparative evaluation, and policy-facing communication through the UBC MDS capstone.

2021 · 2022

Softology IT

Co-Founder / Data Scientist

Client-facing software delivery across applications, APIs, analytics, data workflows, project ownership, and technical handoff.

02 · Engineering range

More than models. The surrounding system matters.

The profile spans AI behavior, machine learning, software interfaces, data foundations, and delivery quality.

01 / AI

Applied AI systems

LLM integration, RAG, agent and tool workflows, structured outputs, prompting, source grounding, human review, and evaluation.

02 / ML

Machine learning

Feature engineering, anomaly detection, ranking, calibration, temporal modeling, clustering, NLP, and reviewer feedback.

03 / SWE

Software & data

Python, FastAPI, REST and WebSocket APIs, SQL, PostgreSQL, BigQuery, pipelines, data contracts, and user-facing integration.

04 / QUALITY

Delivery & trust

MLOps practices, Docker, cloud deployment, CI, automated tests, observability, validation, approval boundaries, and technical documentation.

  1. Data foundationsources, contracts, features
  2. Model or retrievalranking, RAG, prediction
  3. Evaluationquality, failure analysis
  4. Software integrationAPIs, tools, interfaces
  5. Delivery looptests, monitoring, review

03 · Selected work

Public systems for inspecting the engineering.

Two focused case studies provide technical proof without taking over the professional narrative.

Public demo Applied AI Builder 2026

Maintenance-Eye

Multimodal maintenance copilot with live camera and voice input, nine guarded tool workflows, a FastAPI and WebSocket backend, human approval, cloud packaging, and automated tests.

Repository Case study
Public repository RAG system Builder 2026

GovIntel

Procurement intelligence system combining asynchronous ingestion, PostgreSQL, hybrid retrieval, reranking, SQL analytics, structured outputs, and fail-closed citation validation.

Repository Case study

05 · Education

Data science and computer science foundations.

Master of Data Science, University of British Columbia

BSc Computer Science & Engineering, American International University-Bangladesh · CGPA 3.91/4.00

View the full experience profile →

Contact

Building AI or ML systems that need engineering depth?

Start a conversation →