Subject to: Erling D. Andersen

2025-07-31 · Source: Subject to · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Advanced, extended

Summary

Erling Anderson, co-founder and CEO of the Danish company Mosak APS, discusses his journey from a farm in Denmark to becoming a leading developer of optimization software. Born in 1966, Anderson's early life included farming and an introduction to programming in 1981-1983 using Sedex 80/81 computers. He pursued a degree in business administration and economics, where he rediscovered linear programming. His master's thesis focused on implementing the primal-dual interior point algorithm. After completing his PhD in 1992, which included significant work on pre-solving and basis identification, he transitioned to conic optimization during a postdoc in the Netherlands. This led to the creation of Mosak optimization software, officially launched on April 24, 1999, which now serves a diverse global customer base, with 60% of its revenue from financial institutions.

Key takeaway

For optimization software developers and researchers, focus on building robust, theoretically grounded solutions like Mosak's conic optimizer. Prioritize continuous performance improvements for existing features, as this significantly benefits your customer base. Consider offering academic licenses to cultivate a broad user community and identify future talent, and ensure your team structure includes redundancies to maintain long-term project sustainability and knowledge transfer.

Key insights

Mosak's success stems from a deep theoretical foundation, robust conic optimization, and a customer-centric pricing and support strategy.

Principles

• Conic optimization offers superior robustness over general convex solvers.
• Direct internet sales enable competitive pricing and broad reach.
• Building software requires deep theoretical knowledge and early specialization.

Method

Mosak's development strategy prioritizes continuous improvement of existing software for speed and stability, rather than solely chasing novel algorithms, while maintaining a strong academic connection.

In practice

• Utilize conic optimization for robust problem-solving in practice.
• Develop specialized teams with redundancy for long-term sustainability.
• Offer academic licenses to foster community and future talent.

Topics

• Mosak Optimization Software
• Conic Optimization
• Interior Point Methods
• Linear Programming
• Operations Research

Best for: AI Scientist, Software Engineer, Research Scientist

Related on AIssential

• Subject to: Jorge Nocedal — Nocedal's journey highlights the impact of diverse experiences and persistent exploration in shaping foundational contributions to optimization.
• Controlled Self-Evolution for Algorithmic Code Optimization — Controlled Self-Evolution significantly boosts code optimization efficiency by guiding LLM-based iterative refinement.
• Subject to: Aharon Ben-Tal — Ronnie Ben-Tal's career exemplifies the profound impact of theoretical optimization on real-world engineering and medical challenges.
• Applying a Random-Key Optimizer on Mixed Integer Programs — RKO offers a scalable metaheuristic for complex MIPs by decoupling search from feasibility via custom decoders.
• What Makes an LLM a Good Optimizer? A Trajectory Analysis of LLM-Guided Evolutionary Search — Strong LLM optimizers act as local refiners, making incremental improvements within localized semantic search spaces.
• Multi-Objective Constraint Inference using Inverse reinforcement learning — MOCI infers shared constraints and individual preferences from diverse expert demonstrations using an EM-based approach.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Subject to.