Agentic LLM Pipelines
AgentsMemory systems on LangGraph and LangMem that let agents keep track of long tasks instead of losing the thread halfway through.
Emmanuel
Oreoluwa
product · engineer · builder
I build LLM systems and the infrastructure they run on: outreach engines, agent pipelines, and scrapers that keep working when the sites they hit change.
open to opportunitiesread my writing
Focus
What I build
Distributed Scraping
InfrastructureFull scraping systems: sourcing, extraction, proxy and fingerprint rotation, and recovery when a page breaks or an account gets blocked. I build the agents and the pipeline around them, not just the scraper scripts.
Meeting Assistant
VoiceA meeting assistant that transcribes calls in real time, then enriches and structures the transcript so it's searchable and can trigger follow-ups. Whisper handles the audio; the harder part is everything that happens to the transcript after.
Systems & UX Research
ResearchNotes and experiments on system design, working with AI, and keeping interfaces simple. A running interest, not a single repo.
Experience
Senior AI Engineer
@CanAutomatenow
Multi-channel outreach systems across email, WhatsApp, SMS, and AI voice, plus the sourcing and enrichment pipelines behind them.
Senior Software Engineer
Inbox infrastructure that recovers itself when accounts get blocked, scaled horizontally with proxy rotation and browser fingerprinting.
AI & Product
An agentic meeting assistant with real-time transcription and enrichment pipelines. Scaled the infrastructure about 10x.
Product Engineering
React Native apps, focused on getting users through onboarding and on backend speed.
Financial Systems
Payment gateways running at 99.9% uptime, plus lead-gen scrapers powered by Puppeteer and GPT.
Early Career
Cross-platform mobile fundraising tools, plus the system docs and data models behind them.
Writing