Senior Software EngineerAI/ML, Full-Stack & Distributed Systems

Aldrin Cabiong

I build AI agents, full-stack products, and automated workflows that hold up in production.

Senior software engineer with 11+ years across AI, full-stack development, and automation. I design LangGraph agents and grounded RAG, and I ship the APIs, interfaces, and workflows that make them dependable.

  • Bamban, Tarlac, Philippines
  • Clients worldwide · any time zone
Aldrin Cabiong on a mountain overlook
Open to new opportunities
12:00 PMin the Philippines

Now · Self-employed

Since Aug 2026

Working on my own business idea

I have a business idea of my own, and I'm working on the plan, the market, and the first steps to turn it into a real business.

I'm looking for a real partner to start this business with me. Get in touch.

Impact

Years shipping production software
11+
Answer acceptance with grounded RAG, up from 68%
84%
LLM inference cost after context and routing work
−28%
Availability for Kafka-based core services
99.95%

01Selected work

Products I've built and shipped

01/ 12
AI SystemsFull-StackAutomation

Multi-Agent Client Support Platform

LangGraph agents that route, delegate, and act on business systems safely through MCP.

−28%
Inference cost from context budgets and model routing
p95
Latency and tool failures monitored in production
  • LangGraph
  • LangChain
  • FastMCP
Case study
AI Systems

Evidence-Grounded RAG Assistant

Access-aware retrieval with hybrid search, reranking, and citations, measured on a fixed evaluation set.

68%→84%
Human-rated answer acceptance
Cited
Answers link back to their source evidence
  • LangChain
  • LangGraph
  • Hybrid search
Case study
Full-StackAI Systems

Streaming AI Workspace

React and FastAPI interfaces that stream answers and tool steps live, with agent runs on async workers.

Live
Tokens, tool steps, and citations stream live
Async
Agent runs execute on Redis workers, not web servers
  • React
  • TypeScript
  • FastAPI
Case study
AI SystemsAutomation

LLM Evaluation & Cost-Aware Routing

Evaluation suites and model routing that cut inference cost 28% without lowering quality.

−28%
Inference cost from context budgets and model routing
3
Gates on every change: completion, tools, grounding
  • LangSmith
  • LangGraph
  • Python
Case study
AI SystemsAutomationMachine Learning

Engineering Drawing Review Agent

42 deterministic checks across DXF, vector PDF, and scanned drawings, grounded in ASME Y14.5.

42
Deterministic checks automated per drawing
3
Input formats: DXF, vector PDF, and scans
  • LangGraph
  • Tesseract OCR
  • YOLOv8
Case study
AI SystemsAutomation

CRM Workflow Agents & MCP Connectors

LangGraph agents that research, draft, and update CRM records through permission-scoped MCP connectors.

Scoped
Each connector limited to the objects its workflow needs
Idempotent
CRM writes safe to retry without duplicates
  • LangGraph
  • LangChain
  • FastMCP
Case study
Distributed SystemsFull-Stack

Commerce Platform Modernization

Independently deployable services and Kafka event pipelines sustaining 99.95% availability.

99.95%
Availability sustained for core services
−18%
Cloud infrastructure cost via autoscaling and query tuning
  • Kafka
  • Microservices
  • Versioned APIs
Case study
Machine LearningDistributed Systems

Production ML Classification Pipeline

XGBoost with versioned datasets, MLflow tracking, monitored batch inference, and rollback.

MLflow
Every experiment and model version tracked
Versioned
Datasets for reproducible training
  • XGBoost
  • MLflow
  • Python
Case study
Distributed SystemsAutomation

Cloud Cost & Release Engineering

Autoscaling and query tuning that cut cloud costs 18%, plus release checks for consistent deploys.

−18%
Cloud infrastructure cost via autoscaling and query tuning
99.95%
Availability held for core services during the changes
  • AWS
  • Autoscaling
  • PostgreSQL
Case study
Full-Stack

Merchant Dashboards & Payments

React and TypeScript dashboards, checkout, and subscriptions, with load time cut from 3.2s to 1.4s.

3.2s→1.4s
Median dashboard load time
Auditable
Transaction records for every payment event
  • React
  • TypeScript
  • Node.js
Case study
AutomationFull-Stack

CI/CD & Regression Test Automation

GitHub Actions pipelines with API and UI regression tests, promoting one Docker image to every stage.

Every PR
Runs API integration and UI regression tests
1 image
Same Docker build promoted from staging to production
  • GitHub Actions
  • Docker
  • Node.js
Case study
Full-Stack

Scheduling & Customer Records Suite

Node.js and AngularJS apps for scheduling, customer records, and 40% faster recurring reports.

−40%
Runtime of recurring reporting jobs
3 roles
Staff, supervisor, and administrator access
  • Node.js
  • AngularJS
  • MySQL
Case study

02Expertise

AI, full-stack, and automation, end to end

From the model to the interface to the pipeline that keeps it running, with the reliability work that makes it safe to depend on.

AI Agents & LLM Apps

Agents and retrieval that businesses can trust with real tools and data.

  • LangGraph and LangChain agents with supervisor routing, checkpoints, and human approvals
  • MCP tool integrations with validated schemas, scoped authorization, and audit trails
  • Grounded RAG with hybrid search, reranking, and citations
  • LangSmith evaluation suites, failure-injection tests, and cost-aware model routing

Full-Stack Products

Interfaces and APIs that make complex systems usable and fast.

  • React and TypeScript front ends, including streaming AI interfaces
  • FastAPI, Node.js, and Django services with versioned APIs
  • Dashboard load cut from 3.2s to 1.4s with query tuning and caching

Automation & Workflows

Repetitive work turned into reliable, observable pipelines.

  • Document review automated: 42 deterministic checks with OCR and computer vision
  • Celery and Redis task routing with retries and failure recovery
  • Payment webhooks, reconciliation, and CRM workflows with idempotent actions

Cloud & Distributed Systems

Services and pipelines designed to stay up, recover cleanly, and scale cost-effectively.

  • Microservice modernization with backward-compatible database migrations
  • Kafka event pipelines with schema validation, dead-letter queues, and replay
  • AWS, Terraform, Docker, and CI/CD; cloud costs reduced by 18%

03Experience

11+ years of shipping software

Enterprise apps, commerce platforms, event-driven services, and production AI, across on-site, hybrid, and remote teams.

years in production
11+
roles, freelance to enterprise
5
Download résumé (PDF)
  1. Jan 2024 – Jul 2026Full-time · Hybrid

    Senior Software Engineer, AI Systems · Lightweight Solutions

    Metro Manila, Philippines

    • Architected LangGraph agents for client support with supervisor routing, PostgreSQL checkpoints, and resumable human approvals.
    • Introduced MCP integrations (FastMCP) with validated schemas, scoped authorization, audit trails, and idempotency keys.
    • Raised human-rated answer acceptance from 68% to 84% with evidence-grounded RAG; reduced inference costs by 28%.
    • Delivered streaming React/FastAPI interfaces on Redis, Docker, AWS, and Terraform; mentored five engineers.
    • LangGraph
    • LangChain
    • FastMCP
    • LangSmith
    • FastAPI
    • React
    • PostgreSQL
    • Redis
    • AWS
    • Terraform
  2. Oct 2023 – Nov 2025Contract · Remote

    Senior Software Engineer · Freelance

    Bamban, Tarlac, Philippines

    • Built a LangGraph drawing-review agent automating 42 deterministic checks with Tesseract OCR, YOLOv8, and RAG over ASME Y14.5.
    • Engineered asynchronous AI processing with Celery, Redis, and Docker, including retries and failure recovery.
    • Delivered LangChain/LangGraph agents, MCP connectors, SaaS features, and API integrations for client teams.
    • LangGraph
    • Tesseract OCR
    • YOLOv8
    • RAG
    • Celery
    • Redis
    • Docker
  3. Apr 2021 – Dec 2023Full-time · Remote

    Senior Software Engineer · DXC Technology

    Philippines

    • Modernized a client commerce platform into independently deployable services with versioned APIs.
    • Built Kafka order and inventory workflows with schema validation, dead-letter queues, and replay; 99.95% availability.
    • Productionized an XGBoost pipeline with MLflow; reduced cloud costs by 18%.
    • Kafka
    • Microservices
    • XGBoost
    • MLflow
    • PostgreSQL
    • AWS
  4. Jun 2018 – Mar 2021Full-time · On-site

    Full-Stack Software Engineer · Pointwest

    Metro Manila, Philippines

    • Built merchant dashboards, checkout, and subscriptions in React and TypeScript with Node.js and Django APIs.
    • Implemented idempotent payment webhooks, reconciliation, and inventory sync.
    • Cut median dashboard load time from 3.2s to 1.4s with PostgreSQL tuning and Redis caching.
    • React
    • TypeScript
    • Node.js
    • Django
    • PostgreSQL
    • Redis
  5. Jun 2015 – May 2018Full-time · On-site

    Software Engineer · TwistResources, Inc.

    Clark, Pampanga, Philippines

    • Developed scheduling, customer-record, and reporting apps with Node.js APIs and AngularJS.
    • Reduced recurring reporting job runtime by 40% through MySQL index and query tuning.
    • Introduced Mocha/Chai tests, Jenkins builds, and repeatable Linux deployments.
    • Node.js
    • AngularJS
    • MySQL
    • Mocha/Chai
    • Jenkins
    • Linux
  6. 2011 – 2015

    Bachelor of Science in Computer Science · Tarlac State University

    Tarlac City, Philippines

04Skills

The stack I use in production

The tools I reach for every day, then the full toolkit by area. Pick an area to see what I use there.
  • LangGraph
  • LangChain
  • Python
  • FastAPI
  • TypeScript
  • React
  • Node.js
  • Django
  • PostgreSQL
  • Redis
  • Celery
  • Kafka
  • Docker
  • AWS
  • Terraform
  • GitHub Actions
  • MLflow
  • MySQL
  • JavaScript
  • Jenkins
  • Linux

AI Agents & LLM Apps

Agents that route, delegate, and act on real systems safely, with approvals and audit trails.

  • LangGraph
  • LangChain
  • MCP (FastMCP)
  • Supervisor routing
  • Specialist delegation
  • Typed state & checkpoints
  • Human-in-the-loop approvals
  • Tool schema validation

RAG & Evaluation

Retrieval that cites its sources, measured on fixed evaluation sets before anything ships.

  • RAG
  • Hybrid search
  • Reranking
  • Access-aware retrieval
  • Grounded citations
  • Context engineering
  • Model routing
  • LangSmith
  • Evaluation datasets
  • Regression suites
  • Failure-injection testing

Full-Stack Development

Interfaces and APIs that make complex systems fast and usable, including streaming AI UIs.

  • React
  • TypeScript
  • JavaScript
  • Streaming UIs
  • Python
  • FastAPI
  • Node.js
  • Django
  • AngularJS
  • Versioned APIs
  • Mocha/Chai

Automation & Workflows

Repetitive work turned into reliable, observable pipelines with retries and idempotent actions.

  • Celery
  • Redis queues
  • MCP connectors
  • CRM workflows
  • Document automation
  • Tesseract OCR
  • Payment webhooks
  • Idempotent actions
  • Batch jobs
  • GitHub Actions
  • Jenkins

ML & Computer Vision

Models taken to production with versioned data, tracked experiments, and safe rollback.

  • XGBoost
  • MLflow
  • YOLOv8
  • Versioned datasets
  • Batch inference
  • Model rollback

Cloud, Data & Distributed Systems

Event-driven services and infrastructure designed to stay up and scale cost-effectively.

  • AWS
  • Terraform
  • Docker
  • Kafka
  • Microservices
  • Event replay & DLQs
  • PostgreSQL
  • MySQL
  • Autoscaling
  • Linux
  • p95 latency monitoring

05About

The engineer behind the work

Useful AI needs solid software around it.

I've spent more than a decade building software for real businesses, from scheduling and reporting tools to commerce platforms, event-driven services, and, most recently, production AI systems.

My focus is AI that teams can rely on: agents that call real tools safely, retrieval that cites its evidence, and evaluation that catches regressions before customers do. My full-stack and automation background is why I care as much about the API, the interface, and the workflow around the model as about the model itself.

I work with clients worldwide. I'm comfortable working asynchronously, keep people updated without being asked, and can overlap with any time zone when a conversation needs to happen live.

  • Clients worldwide · any time zone
  • BS Computer Science, Tarlac State University
A climber on a snowy summit at golden hour

Based in

Bamban, Tarlac, Philippines

Working with teams worldwide, across any time zone.

01

Measure before shipping

Fixed evaluation sets and regression suites turn "it seems better" into numbers.

02

Design for failure

Retries, idempotency keys, checkpoints, and dead-letter queues make recovery routine.

03

Make actions auditable

Every tool call and approval leaves a trail that people can review.

04

Spend where it matters

Context budgets, model routing, and autoscaling keep quality high and costs in check.

06Contact

Let's build
something real.

Have an AI, full-stack, or automation project in mind? Tell me what you're working on, where it stands today, and what a good result looks like. I'll reply with honest feedback and a clear next step.

  • Replies within one business day
  • Clients worldwide · any time zone
  • Based in the Philippines · UTC+8

Ways to work together

  • Project work

    A defined build with clear scope, milestones, and a clean handover.

  • Long-term contract

    I join your team as a senior engineer, part-time or full-time.

  • Technical advice

    Architecture and code reviews for AI and production systems.

aldrin.t.ca@gmail.com
Email me

Send a message

Tell me what you're working on

What can I help with?

Only used to reply to you. No newsletters.