Peerapon Wechsuwanmanee

Machine Learning Engineer & AI Engineer

Düsseldorf, Germany

Email Peerapon GitHub LinkedIn

Experience

  1. Sep 2024 – Present

    Machine Learning Engineer

    REWE Group · Cologne, Germany

    • Spearheaded the multi-agent AI system that won the company Hackathon 2025 ('Technology Love' category, among hundreds of developers), auto-resolving common software-development issues by surfacing internal documentation and REWE-specific infrastructure fixes - targeting the ~5% of weekly developer time lost to manual lookups and cross-team queries.
    • Architected a multi-agent extension to the store-employee chatbot - natural-language task creation via an MCP tool over a Google ADK agent-to-agent architecture - placing 3rd at Google Agent Factory 2026.
    PythonGenerative AIGoogle ADKMulti-Agent SystemsMCP ServerFastMCP

  2. Jul 2022 – Mar 2024

    Data/Machine Learning Engineer

    Data Reply · Munich, Germany

    • Built a reusable AWS MLOps accelerator - Terraform-provisioned SageMaker training and inference templates with MLflow tracking - to cut time-to-PoC on recurring components across client engagements.
    • Instantiated the accelerator for LLM-based PDF document summarization (LangChain), delivering an end-to-end training-to-serving pipeline.
    PythonSpark ScalaPySparkPyTorchLangChainAWS Glue

  3. Sep 2021 – Jun 2022

    AI Engineer

    Botnoi Group · Remote

    • Led Speech-to-Text product team to build speaker diarization, speaker recognition, and speech recognition models.
    • Built end-to-end ASR pipeline (data acquisition → fine-tuning → deployment) reaching 8% Character Error Rate on the in-domain Thai test set versus 11% for Google Speech Recognition - outperforming the incumbent on a low-resource language in target domains.
    PythonPandasNumPyScikit-LearnPyTorchGCP Cloud Run

  4. Aug 2015 – Jul 2021

    Research Scientist

    Steel Institute (IEHK), RWTH Aachen · Aachen, Germany

    • Drove multiscale, scale-bridging finite-element modelling of damage and formability in advanced high-strength and multiphase (dual-phase) steels - coupling micromechanical ABAQUS simulation (ABAQUS Python API) with experimental material characterization.
    • Authored 9 peer-reviewed journal and conference papers (2014-2021) spanning surface-roughness effects on cold formability, hybrid damage mechanics, hole-expansion and edge-crack sensitivity of multiphase steels, and micromechanical plasticity-and-damage initiation (NUMIFORM, IDDRG, Materials & Design, Computational Materials Science).
    PythonMATLABFinite Element AnalysisABAQUSABAQUS Python APIComputational Mechanics
View full CV →

Featured Projects

Server racks threaded with fiber cabling in a data center

Semantic Product Search Engine

  • Researched and evaluated search engine options for a second-hand e-commerce platform, selecting Meilisearch for its hybrid semantic/lexical search, bilingual Thai/English support, and cost profile at scale.
  • Designed and deployed a real-time PostgreSQL → Meilisearch sync pipeline using logical replication and MeiliBridge, enabling sub-second product index updates without changes to the host application.
MeilisearchPostgreSQLRustTypeScriptAxum
Close-up of a V-twin motorcycle engine and exhaust headers

Motorbike Condition Evaluation

  • Built a multi-modal ML regression pipeline fusing 39-dimensional MFCC audio features (extracted from inspection videos via librosa), Thai-language BERT embeddings (WangchanBERTa), and categorical inspection inputs to predict a 0-100 motorcycle condition score.
  • Benchmarked 5 regression algorithms (ElasticNet, Random Forest, Gradient Boosting, SVR, XGBoost) with scikit-learn RandomizedSearchCV hyperparameter tuning; serialized the best-performing pipeline for cloud-hosted inference.
PythonPandasFastAPIScikit-LearnXGBoost
Candlestick price chart with moving-average trend lines

Credit Risk Assessment

  • Performed exploratory analysis and built a reproducible data-cleaning and transformation pipeline on a raw credit dataset.
  • Ran model selection across candidate classifiers with per-segment personalized feature selection to sharpen credit-default discrimination.
PythonPandasScikit-LearnStreamlit
Macro photograph of a circuit board with gold traces

Credit Default Prediction

  • Built a full-lifecycle credit-default MLOps system: EDA, automated data-cleaning pipeline, MLflow experiment tracking, blue-green model selection, and Prefect-orchestrated automatic retraining.
  • Ran a profit-optimized threshold simulation achieving a 30% profit increase on out-of-sample data versus a baseline cut-off policy.
PythonPandasScikit-LearnFastAPIMLFlow

Skills

Generative AI & Agents

Generative AIAgentic EngineeringMulti-Agent SystemsA2AGoogle ADK

Machine Learning

PyTorchScikit-LearnXGBoostMLFlowPandas

Languages & Frameworks

PythonFastAPIRustAxumTypeScript

Data & Observability

PostgreSQLPrefectGrafanaPrometheusLoki

Cloud, DevOps & Tools

GCPDockerKubernetesTerraformGitHub Actions

Engineering & Simulation

Computational MechanicsFinite Element AnalysisABAQUSABAQUS Python APIMATLAB

Certifications

GCP Associate Cloud Engineer

Google Cloud · 2025

Azure Data Engineer Associate

Microsoft · 2024

Kubernetes Application Developer (CKAD)

CNCF · 2024

Terraform Associate

HashiCorp · 2023

AWS Machine Learning Specialty

Amazon Web Services · 2023

AWS Solution Architect Associate

Amazon Web Services · 2023

Data Scientist Nanodegree

Udacity · 2019

Machine Learning Engineer Nanodegree

Udacity · 2018

From the Blog

All posts →

Credit Risk Assessment When the Data Was Never Built for It

Machine LearningCredit Risk

Scoring Motorcycle Condition from Inspection Videos

Machine LearningMulti-modal

A Full-Lifecycle MLOps System for Credit Default, Part 2: Why Accuracy Is the Wrong Goal for a Lender

MLOpsMachine Learning

Let's talk

Open to machine learning and AI engineering work. Email is the fastest way to reach me.

Email Peerapon LinkedIn