Location
Helsinki, Finland
Ikram Ul Haq
I'm an
ML Engineer II at Smartly in Helsinki, building AI-powered products. Previously led GenAI platform development at Nokia — LLMs, RAG, agentic workflows, and Kubernetes-native microservices with observability-first design.
Get to Know Me
Building AI Products That Ship to Production
ML Engineer II at Smartly in Helsinki, with 6+ years of experience in machine learning and software engineering. Previously an AI Tech Lead at Nokia, where I led a team building a production GenAI platform — RAG pipelines, agentic workflows, and distributed LLM training on private Kubernetes clusters, with observability-first design using MLFlow, OpenTelemetry, and Grafana.
I hold a fully funded Erasmus Mundus dual Master’s in Engineering of Data-Intensive Intelligent Software Systems (EDISS) — one of 23 selected worldwide — from Åbo Akademi University (Finland) and UIB (Spain), specialising in computer vision. I enjoy mentoring engineers, driving technical strategy, and shipping systems built to last.
I love meeting new people — feel free to say hello!
Companies Worked At
ML/GenAI Projects
Years Experience
Programming Competitions Won
Skills
Technical skills developed across production AI systems, cloud-native infrastructure, and NLP research.
Machine Learning & AI
Infrastructure & MLOps
Resume
Projects
A selection of work across GenAI platforms, LLM research, and NLP engineering.
Nokia · 2023 – 2026
Modular GenAI Platform
Led end-to-end development of a production GenAI platform with RAG pipelines, agentic workflows, and distributed LLM training deployed on private on-prem Kubernetes clusters.
Nokia · 2024
Open LLM Evaluation Study
Systematic evaluation of Llama 3, Mistral, and Gemma across summarization, code generation, and conversational tasks. Benchmarked using ROUGE, BLEU, BERTScore, and LLM-as-judge — surfacing key accuracy and consistency trade-offs to guide model selection.
Arbisoft · 2020 – 2022
NLP Review Intelligence
Built NLP tagging pipelines to classify and filter user reviews on travel platforms including KAYAK, enabling hotel-specific tag extraction (cleanliness, service) for smarter personalised search and improved SEO discoverability.
Open Source
Saalis — AI Agent Governance
Built a governance and evidence layer for multi-agent AI systems. Handles output conflicts between agents using arbitration strategies (weighted voting, LLM-as-judge, human deferral), policy enforcement, explainability, and audit logging.
Open Source Contribution · Merged 2025
LEPOR Metric — NLTK
Contributed the LEPOR machine translation evaluation metric to NLTK — one of the most widely used NLP libraries in Python. Implemented sentence_lepor and corpus_lepor with customisable tokenisation options.
Services
Areas where I can help you build, evaluate, and scale AI systems from prototype to production.
RAG & LLM Applications
Designing and deploying Retrieval-Augmented Generation pipelines and conversational AI systems for enterprise search and knowledge management.
Agentic Workflows
Building multi-tool and multi-agent orchestration systems for complex, autonomous enterprise tasks.
MLOps & Observability
Full experiment tracking with MLFlow, distributed tracing with OpenTelemetry, and dashboards via Prometheus and Grafana.
Kubernetes & Microservices
Cloud-native microservice architecture, containerisation with Docker, and on-prem Kubernetes cluster management.
LLM Evaluation & Benchmarking
Rigorous model assessment across summarization, code generation, and conversational tasks — guiding model selection for production use.
Team Leadership & Mentoring
Leading engineering teams, running agile sprints, mentoring junior engineers, and aligning technical roadmaps with business strategy.
Contact
Interested in collaborating on AI products, GenAI systems, or research? Let’s talk.
Contact Info
Based in Helsinki. Open to consulting, collaboration, and interesting AI challenges.
ulhaqi12@gmail.com