I've shipped two live AI products from scratch: CareerLantern, an AI career platform, and Nora, an AI companion for expecting mothers. In parallel, I've spent 5+ years as a PM at Bluewave and TELUS across telecom and AI products. Before product, I spent nearly 3 years building data platforms at Huawei as a software engineer. That means I can read a system architecture, challenge a technical tradeoff, and write a spec that doesn't need translation.
Across these roles, I've increased user adoption by 38%, improved onboarding conversion by 20%, and cut issue resolution time by 25%. I work closely with engineering on data platforms, analytics systems, and AI products, and I'm comfortable getting into the technical details when it matters.
I'm comfortable leading cross-functional teams in Agile environments and translating product vision into outcomes that engineering, design, and stakeholders can all rally around.
Bluewave Mobility Innovation
Product Manager
TELUS
AI Product Manager
Huawei Technologies
Product Manager
Huawei Technologies
Junior Software Engineer
Before moving into product, I spent nearly 3 years as a software engineer at Huawei, building data and analytics platforms. That foundation directly shapes how I work with engineering today. I can read a system architecture, challenge a technical tradeoff, and write a spec that doesn't require translation.
In my own projects, I go further. CareerLantern, a live AI platform I designed and built, gives the clearest view of how I think technically as a PM:
LLM Selection & Migration
Started with self-hosted Llama 3.2 3B via Ollama (zero per-call cost, full control). Migrating to Gemini 2.5 Flash for higher output quality and production reliability. It's a deliberate tradeoff between cost control and model capability as the product matures.
AI Pipeline Architecture
Designed the full LLM pipeline: FastAPI → HTTPS → Cloudflare Tunnel → Ollama. CV parse pipeline: PDF/DOCX → text extraction → prompt (12k char cap) → structured JSON → frontend state. STT pipeline: MediaRecorder → ElevenLabs scribe_v1 → transcribed text.
AI Safety & Guardrails
For Nora (health-adjacent AI), enforced hard system prompt constraints blocking medical advice, with mandatory "consult your doctor" reminders on all AI responses. That was a deliberate product decision balancing user trust, liability, and scope. For CareerLantern, defined output discipline rules (precision, ATS calibration, strict JSON) to ensure consistent, high-quality AI responses at scale.
API Design
Designed and shipped 12 AI endpoints, 7 job tracker endpoints, and a full auth API (JWT + OAuth 2.0). Enforced auth on all AI routes, rate limiting (slowapi), request body caps, CORS lockdown, and disabled API docs in production.
Data & Analytics
At Bluewave, defined KPIs and delivered reporting tools that reduced RCA time by 25%. At TELUS, used data analysis and experimentation to improve chatbot resolution rates. Built dual analytics instrumentation (Mixpanel + PostHog) with a unified wrapper in CareerLantern.
Infrastructure Decisions
Chose DigitalOcean Droplet + Nginx (full VPS control, streaming support) over managed platforms. Private ACL on object storage with presigned URL downloads. APScheduler for scheduled background jobs. SQLite in dev → PostgreSQL in prod with Alembic migrations.
Graduate Diploma in Information Technology
British Computer Society, UK
Bachelor of Science in Physics
Delta State University, Nigeria