In this paper we describe Task-Based Problem Solving Adapters (TPSAs) for modeling a humam solution (through activity-centered analysis) to a software solution (in form of computer-based artifact). TPSAs are derived from the problem solving pattern or consistent problem solving structures/strategies employed by practitioners while designing solutions to complex problems. The adapters developed by us lead toward human-centeredness in their design and underpinning that help us to address the pragmatic task constraints through a range of technologies like neural networks, fuzzy logic, and genetic algorithms. We also outline an example of applying the TPSAs to develop a working system for assisting sales engineers of an electrical manufacturing firm in preparing indent and monitoring the status of orders in the company.