available for architecture & engineering leadership roles

Ponnilatha Ponnaiah Thangaraja

Principal Software Architect

16+ years building distributed systems, event-driven pipelines, and secure microservice platforms for financial services and global logistics — from hands-on Java services at Equifax to architecting cloud-native, authenticated microservice systems at scale, with a growing line of independent AI-native projects.

Apex, North Carolina, USA ponnilatha@gmail.com 972-951-8153 linkedin.com/in/ponnilatha us.theminilatha.com
16+ yrs production systems 6 roles across 5 companies 2 independent AI pipelines shipped 35–50% typical throughput / cycle-time gains
01

Career pipeline

click a node
02

Core competencies

Distributed systems & cloud-native architecture Microservices auth: API Gateway, BFF, service mesh/mTLS, RBAC/ABAC Legacy modernization & system decomposition Event-driven platforms — Kafka, Redis, Kubernetes SOA — Axis2 web services, Java Search & ranking systems Rate management & auditing pipelines AI-native multi-agent pipelines CI/CD, DevSecOps & reliability engineering Architecture governance & engineering leadership
03

Featured personal projects

Independent — decision intelligence

FlightIQ

An AI-powered flight congestion intelligence dashboard that optimizes for outcome, not just price. Surfaces FAA ground-stop programs, airport congestion scores, and 7-day forecasts, then ranks departure windows using graph optimization across cost, time, and congestion. Falls back gracefully to BTS statistical models when live data is unavailable.

NetworkX multi-strategy routing FAA / BTS data
Read full case study →
Independent — AI-native pipeline

Rate Ingestion

An 8-agent AI-native pipeline that extracts structured rate data from PDFs, scanned images, Excel files, and HTML tables. Built for full auditability back to source, cost-aware LLM routing, and deterministic, reproducible output.

multi-agent orchestration LLM extraction cost-aware inference
Read full case study →
Independent — simulation & ops research

Truck & Dock Optimization

A simulation engine modeling how trucks arrive, queue, and get assigned to docks across a working day — realistic arrival distributions, service-time variability, and constraints like equipment type, dock capability, and priority loads. Used to test real-time dock-assignment strategies and see how small logic changes move congestion and idle time. Heading toward a lightweight planning tool for yard/dock managers: live truck-to-dock recommendations, early bottleneck detection, and "what-if" scenario testing without touching live operations.

discrete-event simulation queueing & assignment heuristics AI-assisted scenario generation
Read full case study →
04

Testimonial

"Latha is the kind of engineer that every product owner hopes for: insightful, proactive, pragmatic, and deeply invested in building the right solution within the ever-present time constraints. She was a trusted thought partner whose perspective made our product and our team stronger."

MB
Meghan Barni
Former Product Manager, CargoSphere (dba WiseTech Global)
↓ Read full letter

I am delighted to write this letter of reference for Ponnilatha Thangaraja (Latha). I had the privilege of working closely with Latha for about 6 years while I was a Product Manager at CargoSphere dba WiseTech Global. During that time I was consistently impressed by her technical expertise, leadership, and professionalism.

Latha is an incredibly talented software engineer. She has an impressive ability to combine deep technical knowledge with sound judgement and strong leadership skills. She delivers comprehensive, high-quality solutions to complex problems while keeping the end user and the broader business objectives in mind. She is someone I trusted completely to tackle our most difficult problems.

Beyond her technical abilities, Latha is an outstanding team leader. People naturally look to her for guidance, not just because of her title, but because of her expertise, kindness, dependability, and willingness to mentor. She directs her team with respect and confidence and keeps them focused on the right things at the right times to deliver results.

As a Product Manager, I valued our partnership tremendously. Latha is the kind of engineer that every product owner hopes for: insightful, proactive, pragmatic, and deeply invested in building the right solution within the ever-present time constraints. She was a trusted thought partner whose perspective made our product and our team stronger.

I would enthusiastically welcome an opportunity to work with Latha again and recommend her without hesitation. Any organization would be fortunate to have her, and I am confident she will make a meaningful impact wherever she goes next.

— Meghan Barni, Former Product Manager, CargoSphere

05

Project insights — truck & dock simulation

live model

A simplified version of the underlying queueing mechanics: trucks arrive at random (Poisson process), each dock serves one truck at a time with variable service duration, and a truck waits for the next free dock. Adjust the inputs to see how throughput, wait time, and dock utilization respond — this is the same dynamic the full simulation explores at much higher fidelity.

06

Technology expertise

Languages

Go, .NET 8, Java, Python

Cloud & platforms

Kubernetes, Docker, Kafka, Redis, BigQuery, PostgreSQL

AI & automation

GitHub Copilot, LangChain, retrieval-based systems, model-context servers, multi-agent pipelines

Security & identity

SSO, OAuth2/OIDC, API Gateway, Backend-for-Frontend (BFF), service mesh/mTLS, token introspection, RBAC/ABAC

Legacy enterprise

Java SOA, Axis2 web services, Hybris

DevOps

GitHub Actions, Jenkins, Ansible, Azure DevOps