Angelo Lacson

•Built and maintained production ML infrastructure supporting fraud detection, recommendations, search, personalization, and other prediction-driven applications, working across feature computation, materialization, real-time serving, and batch/streaming pipelines using Python, Spark, Kafka, Snowflake, PostgreSQL, Redis, and AWS while processing more than 50M daily events. •Improved low-latency feature retrieval by optimizing batching, caching, timeout handling, and serving-path behavior, increasing throughput by approximately 2.5x while maintaining sub-100 ms p95 latency for time-sensitive ML workloads. •Designed and implemented a five-agent AI outbound automation pipeline using GPT-4o, CrewAI, Python, Apollo.io, HubSpot, n8n, and Instantly, coordinating Researcher, Analyst, Copywriter, SDR, and CRM agents across prospect discovery, enrichment, personalized outreach, reply handling, meeting workflows, and CRM updates. •Automated prospect research and sales operations using ICP-based sourcing, company signals, LLM-driven personalization, reply classification, and CRM orchestration, reducing research and administrative work from roughly 80% to 20% of rep time and helping increase outbound pipeline volume approximately 3x without additional headcount. •Built AI-assisted enrichment and decision workflows across multiple external data providers and CRM systems, combining LLM-based classification with fallback routing, validation, deduplication, retries, and structured failure handling when upstream data was incomplete or inconsistent. •Engineered synchronization workflows across Stripe and Sage using durable external-ID mappings, idempotent processing, replay protection, reconciliation, and loop-prevention logic for customers, invoices, payments, and accounting-related events. •Developed React-based internal operational dashboards for monitoring AI workflows, enrichment pipelines, synchronization state, failures, and recovery actions, while expanding observability through OpenTelemetry, Prometheus, Grafana, structured logging, distributed tracing, and failure alerts—reducing debugging time by roughly 40%.

Tecton

•Built a production Vendor Risk Analysis AI platform that automated security-document review, compliance mapping, risk analysis, and remediation workflows, reducing vendor-assessment time from most of a working day to approximately 15 minutes per vendor. •Designed and delivered full-stack AI capabilities using Python, FastAPI, TypeScript, React/Next.js, PostgreSQL, and pgvector across document ingestion, semantic chunking, embeddings, vector retrieval, natural-language Q&A, citation mapping, workflow APIs, and internal operational interfaces. •Developed asynchronous document-intelligence pipelines for SOC 2 reports, ISO 27001 evidence, security questionnaires, and penetration-test reports using Redis-backed workers, S3, PostgreSQL/pgvector, structured validation, retries, and failure-safe processing. •Implemented compliance automation across 12 security frameworks and 2,500+ controls, allowing uploaded evidence to be analyzed once and mapped across multiple standards instead of requiring repeated manual review. •Engineered multi-stage full-stack workflow automation for externally integrated applications covering content creation, approvals, scheduling, provider execution, retry handling, reconciliation, and manual-review states using TypeScript, Next.js, Supabase, and third-party APIs. •Designed reliable background and integration workflows using OAuth, atomic work claiming, idempotency, retry/backoff, token lifecycle management, authentication recovery, terminal failure states, and reconciliation, reducing duplicate processing and making external API failures easier to diagnose and recover from. •Built and maintained React/Next.js administration and observability interfaces for monitoring workflow state, provider connections, approvals, failures, scheduled operations, and manual recovery, while hardening services through RBAC, audit logging, rate limiting, secure storage, validation, and automated testing.

Gatekeeper

•Built and shipped customer-facing full-stack products across fintech, healthcare, education, and e-commerce, contributing across React/Next.js frontends, Node.js, Python and PHP/Laravel backend services, REST APIs, relational databases, cloud services, and third-party integrations. •Developed responsive, data-heavy product experiences including financial dashboards, account and transaction workflows, onboarding flows, visual health information, interactive educational interfaces, and customer-facing web applications, translating complex backend data into clear and usable frontend experiences. •Designed backend services and API integrations for payments, customer accounts, notifications, content delivery, authentication, and asynchronous processing, with structured validation, role-based access control, error handling, and maintainable service boundaries. •Built reusable frontend components and application flows with attention to responsive behavior, consistent state handling, form validation, authenticated experiences, API reliability, and maintainability across desktop and mobile web environments. •Owned features through the delivery lifecycle—from requirements analysis and technical design through frontend/backend development, integration, testing, deployment, documentation, client demonstrations, and production handover.

Independent Contributor

AWS Certified Machine Learning – Specialty

AWS

AWS Certified Solutions Architect – Professional

AWS