Tech Lead — AI & Agentic Systems

Amir
Shahbandegan.

I just really love building things.

I'm an engineer who's happiest in the gap between research and production. My favorite work is taking a fresh idea and turning it into a system people can actually rely on. I sweat the details that turn an impressive demo into a tool people can trust.

Start a conversation →

years industry

MSc

computer sci.

∞

curiosity

Toronto

remote / hybrid

§ 01

About

who · what · why

Vector Institute AI SummitToronto · 2025

I build agents that plan, infra that scales, and AI products that ship to production.

I'm currently Tech Lead for AI & Agentic Systems at PwC, leading a cross functional 9-person team building a multi-agent platform on AWS for a Tier-1 Canadian bank which supports hundreds of concurrent long-running agents and powers five finance workflows. Before PwC, I shipped the Canada Revenue Agency's first end-to-end production RAG chatbot, piloted to 500 employees as the foundation for an agency-wide rollout. Earlier, four years as co-founder & CTO of Asan Sports, where I took a sports-booking marketplace from zero to 200K+ users and $1M+ ARR. MSc in CS from Lakehead with ACL-published NLP research.

Agentic AI LangGraph RAG AWS / Azure Kubernetes NLP research CTO / founder

§ 02

Experience

2018 — present · 8+ years

2025 — Present

Tech Lead, AI & Agentic Systems

PwC · Toronto, ON

Leading design, build, and delivery of a use-case-agnostic agentic AI platform on AWS for a Tier-1 Canadian bank. Architected for hundreds of concurrent long-running agents; five finance workflows live. LangGraph orchestration, planner-agent pattern, AES-encrypted state, end-to-end observability. Cross-functional team of nine across five cities.

2024 — 2025

Senior Machine Learning Engineer

Canada Revenue Agency · Remote

Built and shipped the agency's first end-to-end production RAG chatbot on Azure and Red Hat OpenShift. Piloted to 500 internal employees as the foundation for an agency-wide rollout.

2023 — 2024

Machine Learning Engineer

Canada Revenue Agency · Remote

Established MLOps best practices for the AI team. Evaluated tooling and deployment patterns across Azure and AWS. Explored open-source LLMs (Llama, Mistral) for sensitive workloads under federal data-residency requirements.

2021 — 2023

Graduate Research Assistant

Lakehead University — DaTALab · Thunder Bay, ON

NLP and Health AI research, with publications at ACL and other peer-reviewed venues. Fine-tuned LLMs for legal question-answering as part of a Canadian government program.

2022 — 2023

Applied AI Intern

Vector Institute · Toronto, ON

Built Vector Institute's AI self-assessment platform. Designed applied AI programs for industry practitioners and shipped internal automation saving ~360 hours/year.

2018 — 2021

Co-Founder & CTO

Asan Sports · Tehran

Co-founded a sports-booking and coach-marketplace platform. Scaled to 200,000+ users and $1M+ ARR. Owned backend and infra end-to-end — PHP/MySQL core, production on Docker and Kubernetes. Hired and led the engineering team.

§ 03