I build practical Python tools, AI workflows, and connected business systems that reduce manual work and improve everyday operations.
Hi, I’m Lexter. I work with Python, FastAPI, REST APIs, GoHighLevel, n8n, Make.com, and Zapier. From backend applications and web scraping to lead research and CRM automation, I focus on useful, maintainable solutions.
Hi, I’m Lexter. I’m a Technical Virtual Assistant and AI Automation & Backend Developer with a focus on Python, practical software solutions, and organized technical support. I enjoy turning manual tasks into systems that are easier to run and maintain.
My hands-on projects include a multi-provider AI API gateway, a Python lead research dashboard, GoHighLevel lead classification, document processing workflows, and web scraping tools. I have also explored machine learning and desktop application development through personal projects.
I work with FastAPI, REST APIs, JSON, SQL/SQLite, webhooks, GitHub, and automation platforms such as n8n, Make.com, and Zapier. Whether I’m building a tool, investigating a workflow issue, or managing CRM data, I value clear communication, testing, and honest documentation of what works and what still needs improvement.
Python applications, connected APIs, AI workflows, CRM automation, and reliable data support.
Organize contacts, pipelines, custom fields, tags, and CRM workflows.
Keep lead records accurate, stages up to date, and follow-up information organized.
Configure lead routing, follow-ups, reminders, and booking-related automations.
Help with technical tasks, documentation, system checks, and process coordination.
Extract structured information from public web pages using Python and BeautifulSoup.
Discover businesses, review public contact details, and build organized prospect lists.
Normalize, deduplicate, validate, and transform CSV, JSON, and spreadsheet data.
Prepare clean spreadsheets, trackers, and exports for daily operations.
Extract, classify, and route information from supported files and documents.
Store, query, and organize structured records for Python applications and tools.
Write clear setup notes and SOPs so workflows are easier to maintain.
Build task-focused scripts, internal tools, and applications with Python.
Develop API endpoints, backend logic, and integrations for connected systems.
Connect AI models and services to structured application and workflow tasks.
Build and maintain workflows that reduce repetitive work across applications.
Connect apps and services through REST APIs, webhooks, and JSON data flows.
Debug errors, add checks, and document fixes for more reliable workflows.
See my work in Python development, FastAPI, AI integrations, CRM automation, web scraping, and technical experiments.
Built a GoHighLevel-connected workflow that classifies incoming lead messages as hot, warm, or cold. Unclear cases go to manual review instead of being forced into a category. CRM tags and triggers support lead assignment and follow-up.
Watch demoBuilt a FastAPI-based Workflow Agent that extracts content from supported documents, classifies document types, records structured information, and routes items by priority and confidence. SQLite stores processing history, while an n8n webhook is available for workflow integration. The system is rule-based, not LLM-powered; the n8n connection has not been validated in a live deployment.
Developed an OpenAI-compatible FastAPI gateway connecting OpenAI, Anthropic, OpenRouter, and local Ollama models. Implements policy-aware routing, provider fallback, circuit breakers, Server-Sent Events streaming, scoped API authentication, rate limiting, and Prometheus-compatible monitoring. Validated with 125 automated tests and 89 subtests across Windows and Ubuntu in GitHub Actions; production failover has not been independently validated.
View on GitHubBuilt a Streamlit dashboard that discovers mapped businesses using OpenStreetMap, stores research records in SQLite, handles duplicate leads, and tracks follow-up status. Includes location and industry filters, contact-data completeness scoring, optional limited public-email lookup, analytics, and filtered CSV/Excel export. Outreach and lead qualification remain manual. Source code is private.
Private source · Screenshots available on requestBuilt a local V1 backend prototype for handling missed calls, dormant leads, and incomplete bookings. The system includes GoHighLevel messaging/calendar integration points, Gemini-assisted responses, human handoff, and PayMongo payment flow hooks. Local tests passed, but live provider integrations have not been verified end to end.
Prototype · Real-provider deployment not verifiedBuilt a Python web scraper using Requests and BeautifulSoup for extracting structured data from public pages. Supports configurable CSS selectors, SQLite and JSON storage, hash-based duplicate detection, retries, and logging for repeatable research and monitoring tasks.
Independent project · Source not publicly linkedBuilt a Snake-playing reinforcement learning project using a custom neural network and Deep Q-Network training. Includes prioritized experience replay, model save/load functionality, and a Tkinter interface for observing training. This is a personal experiment, not a benchmarked production AI system.
Independent project · Source not publicly linkedDeveloped an experimental Python voice assistant combining speech recognition, text-to-speech, a local Ollama model, and limited desktop actions such as launching approved apps and taking screenshots. Development is currently paused; the assistant is not a finished product.
Unfinished prototype · Development pausedExplored Windows kernel-mode C++ development, storage-stack concepts, and driver debugging with the Windows Driver Kit and WinDbg. This experimental driver did not successfully load, so it is presented as a learning and debugging exercise rather than a working or production-ready driver.
Experimental · Not successfully loadedThis private Python and Streamlit project lets me discover local businesses, review available contact information, track research records, spot duplicate clues, and export prospect lists. Explore screenshots of the application below.
LeadFlow is a local portfolio project. Contact completeness scores do not indicate buying intent, and outreach remains manual. Source code is private.
I can build and troubleshoot solutions, not just operate the tools.
I look for steps that can be simplified with software or automation.
I can collect, clean, and organize data for useful business workflows.
I check how workflows behave and call out unfinished integrations.
I share progress, blockers, and realistic next steps.
I actively explore new tools while staying clear about my experience.
I learn your goals, existing tools, and the task that needs solving.
We agree on scope, integrations, priorities, and expected results.
I review access, data, technical constraints, and workflow requirements.
I develop the solution, check errors, and test realistic scenarios.
I explain the outcome, document key steps, and identify next steps.
What I build, how I work, and how we can collaborate.
My rate is $8/hr, depending on the scope, responsibilities, and type of work required.
I’m based in the Philippines (PHT, UTC+8). I’m open to full-time or part-time remote work and can coordinate hours with US and other international teams.
My main tools include Python, FastAPI, SQL/SQLite, REST APIs, GitHub, GoHighLevel, n8n, Make.com, Zapier, Streamlit, BeautifulSoup, and LLM integrations.
Both. I’m open to short-term builds, ongoing automation work, and full-time or part-time remote Technical VA and development roles.
That depends on the requirements, integrations, and testing involved. I’ll discuss the scope and a realistic timeline before we start.
Yes. Alongside development and automation, I help with lead research, CRM records, data cleanup, documentation, and troubleshooting.
Whether you need an API, AI integration, web scraping tool, GoHighLevel workflow, or hands-on technical support, tell me what you’re working on and we can discuss a practical approach.