Ina Li
AI Automation · GTM Engineering · Product Management
A bit about me
Former AI Project Manager and Product Manager at the Philippines' leading corporate venture builders and technology solutions company.
I bring customer centricity, product thinking, and systems thinking to GTM Engineering and AI Automation — designing and building systems that turn strategy into revenue pipelines, and workflows that streamline real business operations.
What I've shipped
The challenge
No outbound. No owned pipeline. All new clients came through Upwork or one referral partner — neither scalable, neither predictable.
How I solved it
Built a complete cold email outbound motion from scratch — ICP definition, offer design, and a Python scraper that sourced and filtered ICP-qualified partners from directories. Spun up cold email infrastructure with sustained 95–100% deliverability. Configured CRM and sequences so the founder focuses on closing.
Tools
The challenge
Two bottlenecks. Sales reps were spending 30+ minutes manually auditing each clinic website before outreach. And the highest-converting segment — newly opened clinics — didn't exist in any B2B database, so they couldn't be targeted systematically.
How I solved it
Started upstream. Analyzed demo call transcripts and found that ~30% of converted customers were brand-new clinics — pre-vendor, no incumbent to displace, not being systematically targeted. That insight scoped the build.
Lead Enrichment Pipeline: built an n8n workflow triggered by new clinic entries in their Notion CRM. The workflow looks up the clinic website, and crawls it to extract provider count, booking system in use, and competitor signals — replacing the 30-minute manual research with near-instant enrichment.
Business Registry Sourcing: built a Python scraper across the Alberta, Ontario, and British Columbia business registries to surface newly incorporated clinics, then fed them directly into the enrichment pipeline.
Tools
The challenge
Every creative brief meant manually researching Amazon competitor listings and mining product reviews, then writing it all from scratch. Two hours of founder time per brief. No way to scale.
How I solved it
Built a multi-agent Claude Code plugin with an orchestrator that delegates to nine specialist agents — researchers (keyword, competitor, review mining) equipped with tools to fetch live data from the DataForSEO API, and Amazon design specialists (main image, listing images, A+ content).
Amazon listing best practices and the agency's design conventions are embedded into the agents — so briefs come out ready to hand to the designer and client. 2-hour workflow compressed to ~15 minutes.
Tools
Where I've worked
- Led end-to-end delivery of a RAG-based AI customer support agent for a major hospital handling over 100 patient inquiries/day; managed hospital IT stakeholder alignment from discovery through launch.
- Scoped and shipped Synapse, Yondu's internal AI-powered skills assessment and inventory platform.
- Owned end-to-end SDLC for a Legal Research RAG chatbot serving lawyers and paralegals, leading a cross-functional team of data scientists, solution architects, and software engineers under Agile and hybrid delivery methodologies.
- Led delivery of MemoApp's transition from internal Globe Telecom e-signature tool to a multi-tenant SaaS, commercially launched to SMEs.
- Drove 0-to-1 development of an E-Procurement platform for EastWest Bank, replacing a paper-based branch ordering process with a digital workflow integrated directly into the Filinvest ERP — scaled to 400+ branches nationwide.
- Owned product end-to-end: strategy, discovery, requirements, pilot design, and rollout — personally delivered demos, webinars, and training to branch staff across a 4–8 week rollout.
- Led discovery with pilot customers and translated findings into a prioritized feature backlog that shaped the MVP.
- Built no-code automation across the SME loan application-to-disbursement pipeline in ProcessMaker, eliminating manual handoffs and cutting approval turnaround to 1–2 days.
- Wrote Python scripts to automate daily financial reconciliation across eCommerce partner reports, disbursements, and bank statements — designed as an exception-flagging system that surfaced only discrepancies for human resolution.
- Ran the SME credit assessment function — analyzing bank statements, POS records, and revenue proofs — while maintaining a 0% delinquency rate across the portfolio.