Houston, TX · Open to new opportunities

Nana Appiah

Senior Platform & Cloud Infrastructure Engineer

I build the foundations other engineers ship on — AWS networking, identity, Terraform modules, CI/CD, and observability — and lately, the developer tooling that makes that platform easier to use. That same craft is what I'm bringing to trading technology: platforms and pipelines that take quant research from notebook to production without losing rigor.

View projects Get in touch
200+applications migrated to a new cloud platform
50+applications' connectivity restored post-migration
99.99%platform availability on the re-platforming program
67%reduction in IAM configuration complexity
10M+daily ML inference requests served at Trellix
99.999%uptime on ML model serving at Trellix
50%MTTR reduction on Kafka & Elasticsearch at Trellix
50%less developer time answering on-call questions, via an AI chatbot
01

About

I'm a platform and cloud infrastructure engineer with about seven years of experience across fintech, cybersecurity, and investment management.

My career arc runs from ML platform infrastructure at Vanguard, to cloud consulting and data pipelines at Slalom, to SRE/DevOps for multi-regional Kafka, Elasticsearch, and EKS at Trellix — and now to large-scale cloud re-platforming and platform engineering for a fintech, through Evolv Consulting. Lately I've been building developer-facing tooling: reusable Terraform modules, an AI chatbot, and an MCP server.

My consulting background means I'm comfortable with client stakeholders, ambiguous requirements, and translating business needs into infrastructure that actually ships. Core focus areas: AWS cloud infrastructure, platform engineering, Terraform, Python automation, distributed systems, networking, IAM/identity, CI/CD, observability, and on-call incident response.

Outside work I trade stocks and options and manage my own portfolio. That same instinct is what pulled me toward quantitative finance: a strategy is only as trustworthy as the data, backtests, and infrastructure behind it — and that's where I've spent my career. I'm now building the other half of the picture through the MIT MicroMasters in Finance and the Wall Street Quants program, working hands-on with Python-based research workflows: data manipulation, backtesting, portfolio weighting, and performance evaluation.

Working at that intersection — as a quant developer or front-office engineer — is something I'd love to do. But at my core I'm a software engineer, and that's where I want my career to be: building the platforms, pipelines, and tooling that turn ambitious ideas into production systems, whether that's cloud infrastructure or quant research carried from notebook to production without losing rigor.

LocationHouston, TX
Experience~7 years
EducationB.S. Civil Engineering, Computer Science electives — University of Delaware
02

Experience

May 2024 – Present Houston, TX · Remote

Consultant, Senior Cloud Platform / Infrastructure Engineer

Evolv Consulting — client: SoFi, an all-in-one digital platform for banking, lending, and investing

  • Unified AWS Identity Center Permission Set Terraform module: consolidated 3 separate Terraform operations into 1 atomic unit — a 67% reduction in configuration complexity — eliminating drift risk across 2 AWS organizations. 703 lines of module code plus ~2,500 lines of docs; merged with 5 team approvals after ~4 months of iterative prototyping, including terraform state mv migration of existing resources.
  • Cross-account IAM migration automation (Python + Jinja): auto-generated GitLab repo directories, IAM policies, and Terraform code; OIDC trust policies let new-platform (EKS) workloads assume legacy roles. Supported migration of 200+ applications, including sensitive financial-services workloads (banking, debit card, risk), using ForAllValues:StringEquals conditions for shared multi-app service roles.
  • Post-migration network connectivity (VPC peering + PrivateLink via Terraform/GitLab CI): restored connectivity for 50+ applications at 99.99% platform availability — including cross-team VPC peering with the data platform and PrivateLink to a third-party credit API.
  • AWS Network Firewall buildout from zero: stood up a firewall-rules repo end to end (ADR, CI/CD, state, policies). Built a Python Lambda + EventBridge automation that resolves custom-port FQDNs to IPs (AWS Network Firewall only supports FQDN rules on ports 80/443). Migrated 100+ legacy firewall appliance rules across 4 AWS accounts with a default-deny posture for banking environments and STRICT_ORDER stateful evaluation; pruned rules with the Security team.
  • AI-powered Slack chatbot (Lambda + Amazon Lex + LLM via Bedrock/OpenAI): generative answers with a Slack signature verification layer and sync/async invocation paths. Adopted across the Cloud Infrastructure Platform org (5 teams); cut developer time answering on-call questions by 50%.
  • Built a custom MCP server exposing GitLab docs and infra/IaC processes to AI assistants.
  • CI/CD resilience: caching and fallback strategies eliminated GitHub API rate-limit failures; upgraded tflint v0.23.0 → v0.45.0 (2.5 years of improvements).
  • Supported platform tooling for 5 teams — networking and service discovery (security groups, NLB node ports, connectivity-validation test app) plus CI/CD runner IAM for the artifact repo and runner infrastructure.
  • Also: AWS SCPs while provisioning OUs/accounts; break-glass permission sets for on-call; Kubernetes RBAC; SOC2 evidence gathering for access controls; formal on-call with incident response leadership.
Nov 2022 – Apr 2025 Houston, TX · Remote

Staff Software / DevOps Engineer

Trellix — cybersecurity: threat intelligence and detection

  • ML model serving platform (FastAPI + asyncio): 10M+ daily inference requests at 99.999% uptime for the data science team; async, non-blocking I/O to the model backend.
  • Observability & auto-remediation for 5 multi-regional Kafka and Elasticsearch clusters on AWS EKS (Prometheus, Grafana, Bash) — cut MTTR by 50%.
  • Designed and implemented a Consul service mesh: service discovery, mTLS, and traffic management across 5 multi-regional EKS clusters.
  • Kafka broker configuration, producers/consumers, with Cassandra as the primary sink.
  • Built an AWS CDK Python library standardizing EC2/Lambda creation, adopted by 2 teams.
  • Containerized and migrated 50+ legacy Node.js services to cloud-native environments.
  • ArgoCD deployments across EKS; formal on-call; informal mentorship via code reviews and unblocking teammates.
Jun 2021 – Nov 2022 Philadelphia, PA

Consultant, Cloud Infrastructure Engineer

Slalom — global business and technology consulting firm

  • PennyMac (mortgage lending): Python microservices orchestrated by AWS Step Functions processing 500+ financial documents/day via OCR; S3 ingestion; DynamoDB status tracking with retry/catch for idempotent processing; 99.99% uptime.
  • Brigade Energy: led a 5-engineer team building a zero-to-cloud data pipeline (Lambda/Python, S3, Athena, Glue) in 4 months — doubling ingestion/processing throughput. Integrated 4–5 external data vendor APIs and complex SQL parsing; managed client and vendor stakeholders. The team's work earned an 8-week engagement extension.
Oct 2019 – Jun 2021 Malvern, PA

Software Engineer, AI/ML Platforms / Infrastructure

Vanguard — investment management

  • Built an AWS Kendra search index with Python IaC (Troposphere), integrated with frontend search — reduced client inquiry resolution time 10% while maintaining compliance.
  • Co-built a real-time SageMaker Inference Pipeline template (multi-container pre/post-processing) as part of a cross-functional team; adopted by 4 teams.
  • Self-service CloudFormation generation — cut infrastructure dev time 20% across 10+ teams.
  • Provisioned infrastructure across 5 AWS environments with embedded security/compliance safeguards.
03

Projects

Featured · Solo project
GitHub ↗ PyPI ↗

iacgen — IaC Blueprint Generator

A Python 3.10+ CLI (Typer + Rich), published on PyPI as iacgen, that generates opinionated Terraform blueprints for AWS — VPC, EKS, ALB, Kubernetes Service — from a single command or YAML presets.

  • Pydantic config models and a Jinja2 rendering engine with custom Terraform filters
  • Dependency-validation engine enforcing wiring rules (e.g., EKS requires a VPC)
  • pytest suite with 29+ validator tests and GitHub Actions CI/CD
  • Built with help from AI coding assistants
PythonTerraformTyper PydanticJinja2pytest
Hackathon project

AML Detection Engine

Evolv internal hackathon (2025). On a 2–4 person team, I split work between the Snowflake data layer — schema and detection queries — and the Streamlit dashboard.

SnowflakeStreamlitSQL
Open source
GitHub ↗

payments-legacy

A self-contained, EC2-based payments system: three FastAPI services (payments API, ledger, fraud-check) on their own hosts, discovered by DNS, deployed by GitHub Actions through S3 and SSM — with fault injection and load generation built in.

PythonFastAPITerraform EC2GitHub ActionsDynamoDB
Open source
GitHub ↗

project-omnipresent

A push-based configuration management tool built from scratch in Bash to learn how Ansible works under the hood — agentless, host-node architecture over key-based SSH. A learning experiment, by design.

BashSSHDocker
06

Writing

I write about infrastructure, distributed systems, and Python — latest from Medium.

All articles on Medium ↗
04

Skills

Cloud · AWS

EKSLambdaEventBridge Step FunctionsS3DynamoDB AthenaGlueIAM Identity CenterOrganizations / SCPs VPC peeringPrivateLink Network FirewallCloudWatch SageMakerKendraLex BedrockAuto Scaling

IaC

TerraformCloudFormation AWS CDKTroposphere

Languages

PythonFastAPI asyncioPydantic Jinja2Typer BashSQL

Containers & orchestration

KubernetesDocker ArgoCDConsul service mesh

Streaming & data

KafkaElasticsearch CassandraSnowflake Streamlit

CI/CD & observability

GitLab CIGitHub Actions PrometheusGrafana CloudWatch

Identity & security

IAMOIDC Okta SSOKubernetes RBAC SOC2 evidence support

AI tooling

LLM-powered chatbotCustom MCP server Claude Code & Cursor (daily)

Working familiarity

Ansible HashiCorp Vault Datadog
05

Testimonials

Nana handled client stakeholder conversations with grace, navigated risk well, and stepped outside his comfort zone — AWS infrastructure, complex SQL, and vendor API integrations — on a project far more complex than its initial dashboard scope. The team's work earned an 8-week extension.
Former consulting colleague
"I don't care what we need to do to keep this team on board but let's make that happen."
Client CTO
Nana engaged disparate client team members to refine unclear requirements with little oversight, made solid backend contributions, and kept a positive, learner's attitude and poise under pressure.
Former consulting colleague
07

Contact

The fastest way to reach me is email — I read everything.

ny.nappiah@gmail.com
LinkedIn ↗ GitHub ↗ Medium ↗

Based in Houston, TX · Remote-friendly