Machine Learning Engineer · ML Platforms · Agentic systems · Production Software
@agent(id="fadhilmch", based_in="Stockholm 🇸🇪", roots="Indonesia 🇮🇩")
class Fadhil(MachineLearningEngineer):
"""Ships the model, then keeps it alive."""
toolkit = [Python, Go, JS, K8S, Kubeflow Pipeline, Azure, GCP]
def solve(self, problem):
while not problem.is_clear:
problem = self.ask(problem, "but why?")
return self.ship(simplest_thing_that_works(problem))
@llm_eval(metrics=["relevance", "retrieval_recall", "groundedness", "latency"])
def did_it_work(self, response):
return response.is_relevant and response.is_grounded
def on_error(self, error):
return self.learn(error) # production is where the interesting bugs live- 💻 A Machine Learning Engineer & Software Engineer at
Electrolux - 📚 Amplified my passion in Machine Learning with an MSc from
KTH Royal Institute of Technology, Sweden - 💡 Got my entrepreneurial hat on as the Founder of Icarus Virtual Sports Platform and some small SMEs.
- 🎸 Spent some rad time before as Machine Learning Engineer Intern at
Spotify
- 🤖 Building agentic workflow and multi-agent system for business process improvements.
- 🏗️ Building Experimentation Platform to fuel the data-driven culture & decision-making.
- 🎰 Developing Multi-armed Bandits Model for optimizing ML models deployment & experimentation.
- 🌍 Exploring Kubeflow Pipeline to support ML models lifecycle.
- 🧪 Crafting and deploying machine learning models to support business needs.
- 🏛️ Creating robust software architecture and cloud solutions.
- 💡 Digging deep into data to pull out valuable insights.
- 🚀 Optimizing product development through data-driven experimentation.
- 🌱 Spreading the innovation with the power of AI and data-driven decision-making.
- 🔬 Evaluation that tells us whether an AI system actually works
- 🛠️ ML platforms that make the right path the easy path
- 🧪 Experimentation infrastructure that helps teams learn instead of guess
- 🧱 Boringly reliable services: observable, fault-tolerant, and simple to operate
- 💡 Starting with the user and the problem, then choosing the technology
Python · Go · SQL · Kubernetes · Kubeflow Pipelines · Terraform · Azure · Google Cloud
· BigQuery · Redis · MLflow · RAG · LLM evaluation · A/B testing · multi-armed bandits
If you're working on production AI, ML platforms, experimentation, or evaluation, I'd enjoy the conversation.
