Toronto based and open to hands-on AI roles

AI Product Lead and Solutions Architect

I turn unclear operating problems into working AI products.

I lead the work from discovery and requirements through architecture, product logic, implementation, testing, and release. My background spans banking systems, content moderation tools, manufacturing quality, and hands-on AI product delivery.

Mohammed Omer Shaik

Selected work

Products with the evidence left visible

Each product is labelled by who it was built for, what reached a working state, and where the boundary still sits.

Client product

SANAD

Live demo
Preparing for commercial launch

The problem: contractor discovery and procurement depend on fragmented records, unverified claims, and bid processes that are difficult to compare. I built a bilingual product for a founder that brings discovery, verified profiles, project posting, comparable bids, awards, milestones, purchase orders, invoices, and audit records into one workflow.

126product screens
452registered states
904English and Arabic browser cases
EN + ARfull RTL support
SANAD contractor discovery home screenSANAD contractor profile with credentials and trust score

Current boundary: the product and bilingual state coverage are working. External verification, messaging, and payment services remain behind replaceable adapters for production integration.

Open live demoSource repository is private

Partner backed product

LaserEyedBen

Production system
Paused after operation

An autonomous crypto intelligence agent that combined market, on-chain, developer, news, and social signals. I designed the multi-source workflow, event processing, retrieval, personality controls, fallbacks, rate limits, and operating checks.

~16integrated data sources
20+intelligence engines
85%internal cache hit benchmark
<1.5 sinternal response benchmark

Current boundary: the system ran publicly and was paused when API and infrastructure costs outgrew available funding.

Studio product

Crucible

Working prototype
Evaluation system

A product specification review system that finds missing acceptance criteria, conflicts, unsupported claims, hidden assumptions, and decisions that still require human judgment. The readiness score cannot show complete while a conflict or judgment item remains open.

Crucible review canvas showing readiness, open decisions, conflicts, and inline findings

The prototype includes a typed rubric engine, contextual grounding, lossless export, API validation, and automated tests. The visible demo uses a deterministic stand-in when no model key is present.

Repository available on request

Studio product

Lorenth

MVP
Agentic research

A research system that turns a question into a structured, sourced report. It plans the work, gathers current evidence, synthesizes cited findings, and exports the result as a document, presentation, or PDF.

Lorenth report export template controls

Built with FastAPI, React, PostgreSQL, Redis, Claude, Perplexity, live data connectors, authentication, billing, and document export services.

Studio product

Conviction

Live beta
Decision system

A decision-quality journal for thesis-driven investors. A thesis cannot save without a falsifier, versions remain immutable, reality checks flag broken assumptions, and a deterministic process score evaluates the quality of the decision rather than the return.

Conviction decision journal title card

No recommendations, broker connections, or automated trading. The product is deliberately limited to recording, checking, and reviewing a user's own reasoning.

Open live betaSource repository is private

Approach

Product judgment backed by systems work

I started by building and operating a local commerce product, moved into campaign operations, then spent four years translating banking workflows into services, APIs, data models, and release requirements for engagements serving HSBC Bank and Saudi National Bank.

That path matters because AI products fail for ordinary reasons: unclear ownership, weak data, missing states, unsafe actions, and no plan for failure. I work across the product and technical layers so those questions are answered before a demo becomes an operating problem.

For high-risk workflows, I add citations, validation, permissions, audit records, fallbacks, and human approval. The goal is a useful system that people can understand and operate, not a black box with an impressive prompt.

01 · Frame

Find the decision and the user

Map the workflow, source material, constraints, failure cost, and the smallest release worth testing.

02 · Build

Connect product logic to the system

Write the requirements, architecture, data contracts, interfaces, evaluation rules, and working implementation.

03 · Prove

Test the states, not just the happy path

Check failure behavior, permissions, evidence, release controls, and the boundary between working and planned.

Product

Discovery, workflow mapping, PRDs, user stories, acceptance criteria, roadmaps, backlog decisions, demonstrations, release planning

AI systems

LLM APIs, agents, retrieval, structured outputs, evaluation, prompt and context design, citations, fallbacks, human review

Software and data

Python, TypeScript, JavaScript, Java, React, Next.js, FastAPI, Node.js, PostgreSQL, Redis, Supabase, vector retrieval

Experience

A clean path from operations to AI product delivery

The chronology here matches my résumé.

AI Product Lead and Solutions Architect at CodeSpellStudio

Canada and India, remote

Led discovery, requirements, architecture, implementation, testing, release controls, demonstrations, and documentation across client engagements, partner backed products, studio products, and open source work.

Product Owner, Content Moderation Tools at Majorel Canada

Canada

Turned recurring reviewer, policy, privacy, quality, and escalation issues into clearer workflow requirements and recommendations, working across operations, quality, engineering, and linguistics.

Quality Assurance and Process Improvement Analyst, Co-op at Elite Mobility Mfg.

Mississauga, Ontario

Analyzed manufacturing and quality data, supported statistical process control and model validation, and streamlined inspection work to reduce quality check time by 25 percent within two months.

Software Architect and Technical Product Lead at Bismi Services Pvt. Ltd.

Hyderabad, India · Banking engagements for HSBC Bank and Saudi National Bank

Translated banking operations and data requirements into service boundaries, API contracts, relational schemas, integration logic, reporting workflows, acceptance criteria, and production support decisions.

Ad Operations Specialist, Campaign Delivery and Quality at Yahoo

India

Supported digital campaign setup, delivery monitoring, data quality, reporting, and issue resolution, with careful checks across targeting, creative, tracking, and delivery records.

Founder, Product Manager and Operations Lead at Instazop

Kalaburagi, India

Built an early online grocery marketplace from concept through launch, managed vendor and delivery operations, supervised eight delivery personnel, and served an average of 20 customers a day in the first month.

Education and credentials

Technical foundation and product practice

Ontario Graduate Certificate

Artificial Intelligence and Machine Learning

Lambton College, Toronto · Jan 2022 to Aug 2023

Bachelor of Engineering

Computer Science and Engineering

P.D.A. College of Engineering · Completed Jul 2017

Certifications

Professional Scrum Product Owner I, Scrum.orgAI in Fintech Essential Training, LinkedIn Learning, 2023Excel Essential Training (Microsoft 365), NASBA, Jul 2023

Contact

Have an operating problem worth solving?

I am open to AI product, technical product, solutions architecture, applied AI, and automation roles.