Software Engineering & Leadership

A showcase of my software engineering and leadership journey, projects, and professional experience.

Professional Experience

Senior Backend Engineer at Tibo Energy

2025-02 – Present

Designing, implementing, and maintaining backend features and services for an industrial energy-management platform, building service-layer logic and data models across a TypeScript and Node.js stack backed by MySQL and Cassandra.

Technologies used:

  • TypeScript
  • Node.js
  • MySQL
  • Cassandra
  • REST APIs

Chief Technology Officer (Part-time) at BrainBite.ai

2024-12 – 2025-09

Defined the company's technology goals and owned the architecture decisions for its AI vertical on an AI-powered personalised learning platform. Led reliability improvements across LLM-based content-generation modules, and established both automatic and manual evaluation methodologies, including LLM-as-a-judge and red-teaming, to measure and improve the quality and safety of the AI systems.

Technologies used:

  • Python
  • Large Language Models
  • RAG
  • LLM-as-a-Judge
  • Red Teaming
  • Evaluation Design
  • System Architecture

Fractional CTO (Part-time) at ad:personam GmbH

2021-01 – Present

Built the first MVP and designed and developed the web application for a self-serve programmatic-advertising Demand-Side Platform (DSP). Delivered the backend end-to-end, including third-party integrations (Xandr, Stripe, OpenAI), CI/CD pipelines, and AWS VPC-based security, and later led the migration of the infrastructure to Microsoft Azure. Owned infrastructure policy, security and privacy, and cost management, and guided the hiring and management of technical staff.

Technologies used:

  • Java (8/17)
  • Spring Boot
  • MySQL
  • AWS (VPC, Lambda, API Gateway, Cognito)
  • Microsoft Azure
  • OpenAI / RAG
  • REST APIs
  • CI/CD

Doctoral Researcher at University of Twente

2020-09 – Present (full-time to 2024, part-time since)

Studied how conversational agents can adapt their language to individual users, through lexical alignment, to improve trust, comprehension, and persuasion, combining controlled user experiments with computational dialogue modelling and grounded in psycholinguistics and cognitive science. Designed and built conversational agents that generate interactive, personalised explanations, and developed weighted-decoding methods for controllable lexical alignment with large language models such as Llama, Phi, and BlenderBot (under review at SigDial 2026). Ran empirical, between-subjects user studies measuring trust, persuasiveness, cognitive load, and information uptake. Supervised eight MSc and BSc students, and published at ACM IUI, ACM UMAP, and Computer Speech & Language. Marie Skłodowska-Curie fellow under the EU-funded NL4XAI project.

Technologies used:

  • Python
  • NLP
  • Large Language Models (Llama, Phi, BlenderBot)
  • Weighted Decoding
  • Rasa
  • PyTorch
  • spaCy
  • Experiment Design
  • Statistical Analysis
  • Scientific Writing

MSc Researcher in Artificial Intelligence at KU Leuven

2019-09 – 2020-05

Researched the automatic interpretation of simple metaphors, modelling pragmatic inference with a Bayesian Rational Speech Act framework over ConceptNet as a large-scale knowledge graph. Proposed and developed the research topic independently. Graduated Cum Laude.

Technologies used:

  • Python
  • Bayesian Inference
  • Rational Speech Act
  • NLP
  • ConceptNet
  • Knowledge Graphs

Team Manager & Technical Lead at MiQ Digital

2017-01 – 2019-09

Led and mentored a team of seven engineers and owned seven production services. Re-architected a real-time ad-bidding platform to handle 1.5M requests per second at sub-2ms p99 latency, using an event-driven architecture and database optimisations. Defined company-wide code and documentation standards, and contributed performance improvements to an open-source PostgreSQL library.

Technologies used:

  • Java
  • Vert.x
  • PostgreSQL
  • AWS
  • Nginx
  • Event-Driven Architecture

Senior Software Engineer at MiQ Digital

2015-01 – 2016-12

Developed a real-time bid-modification service processing up to 1.5M requests per second. Modernised 70% of a core interface service through refactoring, library upgrades, and improved modularity. Built a horizontally scalable pipeline to compress and ship up to 90 TB of logs per day to AWS S3.

Technologies used:

  • Java
  • AWS S3
  • PostgreSQL
  • Distributed Systems

Software Engineer at MiQ Digital

2013-09 – 2014-12

Designed a data-ingestion service supporting 12+ protocols and 30+ heterogeneous data sources, with custom processing pipelines persisting data to AWS S3 and Redshift. Mentored new engineers and contributors.

Technologies used:

  • Java
  • AWS S3
  • AWS Redshift
  • Data Pipelines

Software Engineer at Apostek Softwares Pvt. Ltd.

2012-07 – 2013-08

Built and maintained a mobile advertising platform and a points-based rewards system, in which users earn points for watching ads and redeem them for rewards. Delivered critical enhancements, bug fixes, and new features across multiple portals using Ruby on Rails with Cassandra and MySQL, and HTML, CSS, and JavaScript for the interfaces.

Technologies used:

  • Ruby on Rails
  • Cassandra
  • MySQL
  • JavaScript
  • HTML/CSS

Selected Projects

Conversational Agent for Personalized Explanations

A dialogue system that generates interactive, personalized explanations using lexical alignment techniques. Built as part of PhD research at the University of Twente under the EU-funded NL4XAI project.

Technologies: Python, Rasa, PyTorch, Large Language Models, spaCy

View on GitHub

Real-time Ad Bidding Platform

Revamped a high-performance real-time bidding platform handling 1.5M requests per second with sub-2ms p99 latency using event-driven architecture and database optimizations at MiQ Digital.

Technologies: Java, Vert.x, PostgreSQL, AWS, Nginx

Scalable Log Shipping System

Built a horizontally scalable system to compress and ship up to 90 TB of logs daily to AWS S3, enabling efficient data warehousing and analytics.

Technologies: Java, AWS S3, Distributed Systems

I Think Blah - Personal Website

This website! A Jekyll-based personal site featuring research, software portfolio, philosophy blog, and food recipes.

Technologies: Jekyll, HTML/CSS, JavaScript

View on GitHub

Technical Skills

  • Programming Languages: Java, Python, TypeScript, JavaScript, Ruby
  • AI & NLP: PyTorch, Rasa, spaCy, NLTK, Large Language Models, Google DialogFlow
  • Backend & Web Servers: Vert.x, Node.js, Nginx, Apache Tomcat, Ruby on Rails
  • Databases: PostgreSQL, MySQL, Cassandra, MongoDB, AWS Redshift
  • Cloud & Infrastructure: AWS (S3, Redshift), Google Cloud, Docker, Git, Linux
  • Research: Experiment Design, Statistical Analysis, Scientific Writing, User Studies

For more details about my professional experience, please visit my LinkedIn profile.