Skip to content
FH

Manager, System Integrations · CapTech

Hi, I'm Nando.I build cloud platforms and AI agents on Azure.

Fernando Huamancaja. I design Azure infrastructure in Terraform, orchestrate AI agents in Python with Semantic Kernel on Azure AI Foundry, and ship the APIs and event-driven integrations that connect them to real systems.

LinkedIn GitHub Get in touch
  • Azure
  • Terraform
  • Python
  • Semantic Kernel
  • AI Foundry
  • .NET
Portrait of Fernando Huamancaja
Azure
Semantic Kernel
Scroll to learn more

About

Cloud, AI and the integrations in between.

I am a software engineer turned cloud and AI integration lead. At CapTech I design and ship solutions on Microsoft Azure: infrastructure defined in Terraform and delivered through Azure Pipelines, event-driven integrations built on Service Bus and Azure Storage, and AI workloads running on Azure AI Foundry.

My current focus is agentic AI in Python. I orchestrate multi-agent workflows with Semantic Kernel, expose them as API endpoints, and build scoring and evaluation agents so quality is measured, not assumed.

Before consulting I spent several years in financial services at LPL Financial and Inspira Financial building .NET and Angular platforms, and I started my career in networking and IT support at Lowe's.

Based in
Charlotte, NC · Remote
Currently
Manager, System Integrations, CapTech

What I do now

  • Azure cloud & infrastructure as code

    Terraform modules and environments deployed through Azure DevOps pipelines.

  • AI agent orchestration

    Semantic Kernel agents on Azure AI Foundry, exposed as APIs, with evaluation and scoring agents.

  • Event-driven integrations

    Service Bus messaging, Blob and Table storage, and data indexers feeding search and retrieval.

  • .NET engineering & delivery

    C# services, APIs and CI/CD with a background in .NET migrations and production support.

Experience

Where I've built things.

From network operations to financial services to consulting on Azure and AI.

  1. CapTech

    Technology consulting firm delivering cloud, data and AI solutions for enterprise clients.

    Mar 2025 – Present United States

    Manager, System Integrations

    Mar 2025 – Present
    • Design and build Azure infrastructure as code with Terraform, and deploy it through Azure DevOps pipelines across environments.
    • Develop AI solutions on Azure AI Foundry, orchestrating multi-agent workflows with Semantic Kernel in Python and exposing them as API endpoints.
    • Build scoring and evaluation agents that grade agent and API responses so quality is measured before and after every release.
    • Design event-driven integrations with Azure Service Bus and Azure Storage (Blob containers and Table storage).
    • Build data indexers and ingestion pipelines that feed search and retrieval for AI workloads.
    • Lead technical delivery with client stakeholders and engineers.

  2. Inspira Financial

    Financial services company providing health, wealth, retirement and benefits solutions.

    Nov 2024 – Mar 2025 United States

    Software Engineer

    Nov 2024 – Mar 2025
    • Full-stack .NET and Angular development on financial services platforms.
    • Delivered features in an agile team alongside product, QA and platform engineers.

  3. LPL Financial

    Largest independent broker-dealer in the United States.

    Jun 2021 – Oct 2024 Fort Mill, SC (Remote)

    Senior Software Engineer

    Jun 2023 – Oct 2024
    • Owned features end to end on .NET 6 and Angular applications used by advisors and operations teams.
    • Troubleshot production issues during release nights and drove fixes across teams.
    • Led research and discovery spikes with partner teams and documented recommendations.
    • Created Cucumber acceptance tests with the QA team to harden regression coverage.

    Software Engineer II

    Jun 2021 – Jun 2023
    • Migrated applications from .NET Framework 4 to .NET 6 and introduced the Mediator pattern in the migrated projects.
    • Built controller endpoints and business logic following the Repository pattern.
    • Maintained TeamCity and Octopus CI/CD pipelines and resolved build and deployment issues.
    • Worked with the AWS Management Console and CLI and used GitHub for source control.

  4. CanvasPeople

    E-commerce company selling custom canvas prints.

    Feb 2021 – Mar 2021 Remote

    .NET Engineer (Contract)

    Feb 2021 – Mar 2021
    • Extended a nopCommerce store built on .NET Core MVC, Razor Pages, Entity Framework and MSSQL.
    • Delivered in Scrum sprints with the client's product team.

  5. Lowe's Companies

    Fortune 50 home improvement retailer.

    Mar 2017 – Feb 2021 Mooresville, NC

    Network Engineer / .NET Developer

    Nov 2019 – Feb 2021
    • Created NocBox, a C# .NET WinForms troubleshooting tool backed by SQLite and Dapper that the network operations team used daily.
    • Built ShiftChecker, a Xamarin.Forms mobile app for shift tracking.
    • Troubleshot routers, switches, wireless controllers and access points and drove escalations with higher tiers and ISPs.

    IT Support Analyst

    May 2018 – Nov 2019
    • Subject matter expert for mobility and supply chain systems on the night shift.
    • Built an internal HTML/JavaScript knowledge page and trained new hires.

    Service Desk Technician

    Mar 2017 – May 2018
    • Supported Windows and iOS devices using PuTTY, AirWatch, Active Directory and SCCM remote tools.

Work

Selected work.

Recent client work at CapTech, described at the level of problem, approach and outcome, plus earlier side projects.

AI agent orchestration platform

Lead engineer · CapTech · 2025 – present

Multi-agent workflows on Azure AI Foundry, orchestrated with Semantic Kernel in Python and exposed as API endpoints.

Problem
Business processes needed language-model reasoning over internal documents and systems, with quality that could be proven, not assumed.
Approach
Semantic Kernel agents in Python with plugins for internal APIs, Service Bus for asynchronous hand-offs, Blob and Table storage for state, and indexers that keep retrieval grounded.
Outcome
Dedicated evaluator and scoring agents grade every agent and API response, so regressions surface before release instead of in front of users.
  • Azure AI Foundry
  • Semantic Kernel
  • Python
  • Service Bus
  • Azure Storage

Azure infrastructure as code

Cloud engineer · CapTech · 2025 – present

Reusable Terraform modules and environment definitions for Azure, delivered through Azure DevOps pipelines.

Problem
Environments were being assembled by hand, which made them slow to reproduce and hard to review.
Approach
Terraform modules for networking, compute, storage, messaging and AI services, with remote state and per-environment variable sets.
Outcome
YAML pipelines plan and apply every change with approvals per environment, giving reviewable, repeatable infrastructure from dev to production.
  • Terraform
  • Azure
  • Azure Pipelines
  • YAML

Event-driven system integrations

Integration engineer · CapTech · 2025 – present

Service Bus queues and topics connecting line-of-business systems with resilient, idempotent .NET consumers.

Problem
Point-to-point calls between systems were brittle and failed loudly whenever a downstream service was slow or unavailable.
Approach
Decoupled producers and consumers with Service Bus, retries and dead-letter handling, and ingestion pipelines that index data for search and AI retrieval.
Outcome
Integrations absorb downstream outages, replay safely, and give operators a clear view of every message that needs attention.
  • Azure Service Bus
  • Blob Storage
  • Table Storage
  • Azure AI Search

More projects

Reena's Blog

March 2022

Serverless blog on AWS. Backend with AppSync, Amplify CLI, DynamoDB, Amplify Auth and Cognito; frontend with Next.js and Tailwind CSS.

  • AWS Amplify
  • DynamoDB
  • Next.js

MovieSearch

March 2021

TMDB-backed movie search. ASP.NET Core API with Entity Framework and an Angular client, run against SQLite and a PostgreSQL Docker container.

  • ASP.NET Core
  • Angular
  • PostgreSQL

DatingAppUpdated

January 2021

Full-stack course project on ASP.NET Core 5 and Angular 10 with TypeScript, covering auth, SignalR and paging.

  • ASP.NET Core
  • Angular
  • SignalR

NocBox 2.0

December 2020

C# WinForms troubleshooting tool for a network operations team. Reads CSV, Excel and SQLite data through Dapper and Entity Framework.

  • C#
  • WinForms
  • SQLite

Shift Checker

August 2020

Xamarin.Forms mobile app that tracks shift schedules, days remaining and hours left, with SQLite local storage.

  • C#
  • Xamarin.Forms
  • SQLite

Hospital Management System

September 2020

Employee login and medical records desktop app on .NET Framework with LINQ, Entity Framework and SQLite.

  • C#
  • Entity Framework
  • SQLite

Skills

Tools I reach for.

Grouped by the kind of problem they solve.

Cloud & infrastructure

  • Microsoft Azure

    Compute, storage, messaging, identity and AI services across multiple environments.

    • App Services
    • Functions
    • Key Vault
    • Entra ID

  • Terraform

    Reusable modules for Azure landing zones and application infrastructure, promoted through dev, test and prod.

    • Modules
    • Remote state
    • azurerm

  • Azure DevOps Pipelines

    YAML pipelines that plan and apply Terraform, run tests and deploy .NET services with approvals.

    • YAML pipelines
    • CI/CD
    • Environments

  • Azure Service Bus

    Queue and topic based integrations with retries, dead-lettering and idempotent consumers.

    • Queues
    • Topics
    • Event-driven

  • Azure Storage

    Blob containers for documents and artifacts, Table storage for lightweight state, lifecycle and access policies.

    • Blob
    • Table
    • SAS

  • Docker & Kubernetes

    Containerized .NET services running locally and in the cloud.

    • Containers
    • Compose
    • AKS

AI engineering

  • Azure AI Foundry

    Model deployments, prompts and agents managed on AI Foundry and wired into production .NET applications.

    • Model deployments
    • Agents
    • Prompt flow

  • Python for AI

    The language I build agents, evaluators and data pipelines in, from async services and typed models to the API endpoints around them.

    • Python
    • asyncio
    • Pydantic

  • Semantic Kernel

    Multi-agent workflows in Python with plugins, planners and function calling, exposed through API endpoints.

    • Agents
    • Plugins
    • Function calling

  • Agent evaluation & scoring

    Evaluator agents and scoring pipelines that grade responses for accuracy, groundedness and regressions.

    • Evaluators
    • Scoring
    • Quality gates

  • Search & data indexers

    Document and structured-data ingestion that gives agents grounded context to retrieve.

    • Azure AI Search
    • Indexers
    • RAG

  • Claude & LLM APIs

    Claude Certified Architect (Foundations). Prompt, tool and agent patterns that hold up in production.

    • Anthropic
    • Tool use
    • Prompting

Backend & data

  • C# / .NET

    ASP.NET Core APIs, background services, Mediator and CQRS patterns, and .NET Framework to .NET Core migrations.

    • .NET 8
    • ASP.NET Core
    • MediatR

  • Entity Framework & LINQ

    Code-first and database-first data access, migrations and query tuning.

    • EF Core
    • Dapper

  • SQL Server & PostgreSQL

    Relational modeling, stored procedures and performance work on MSSQL, plus PostgreSQL and SQLite for smaller services.

    • MSSQL
    • PostgreSQL
    • SQLite

  • Testing & quality

    Unit and integration tests in .NET, Cucumber acceptance tests, and production release support.

    • xUnit
    • Cucumber
    • Release nights

Frontend & tooling

  • Angular & TypeScript

    Enterprise Angular front ends paired with .NET APIs, from component libraries to build pipelines.

    • Angular
    • TypeScript
    • RxJS

  • Git & GitHub

    Branching strategies, pull request reviews and GitHub-based delivery workflows.

    • GitHub
    • Azure Repos

  • AWS

    Amplify, AppSync, DynamoDB and Cognito for serverless side projects, plus AWS console and CLI at LPL.

    • Amplify
    • DynamoDB
    • Cognito

  • Linux

    Comfortable on the command line for containers, pipeline agents and troubleshooting.

    • Bash
    • RHEL

Certifications & education

Credentials.

  • Claude Certified Architect – Foundations

    Anthropic · Sep 2026

    First step in the Architect track. Validates the skills to design, build and deliver production-ready solutions with Claude.

    View credential

  • Microsoft Certified: Azure Fundamentals (AZ-900)

    Microsoft · Apr 2021

    Cloud concepts, core Azure services, security, privacy, compliance and pricing.

  • YearOne Mentorship Finalist

    YearOne · Jan 2021

    Selected from nearly 800 candidates as a YearOne programming finalist.

  • Computer Network Integration

    Central Piedmont Community College · Dec 2012

    Graduated with a 4.0 GPA in Charlotte, NC.

Contact

Let's build something on Azure.

Whether it is a Terraform platform, an integration that needs to stop breaking, or an AI agent you want to trust in production, I would enjoy hearing about it.

fahuamancaja@gmail.com LinkedIn