Appendix F — About the Author

---

---

F.1 Profile #

Ammar Jamshed works at the intersection of enterprise data architecture, applied machine learning, and analytics practice. His interest in the problem this whitepaper addresses arose from a pattern encountered repeatedly across data work: organizations that had successfully solved storage, integration, and processing still could not readily answer questions that spanned more than one domain — and rebuilt the same cross-domain reasoning, from scratch, every time it was needed.

The Data Swimming Pool framework is the result of attempting to describe what an architecture would look like if the relationship between data — rather than the data itself — were treated as the asset worth building infrastructure around.

F.2 Areas of Focus #

F.3 Contact #

F.4 An Invitation to Disagree #

The most useful response to this whitepaper would be a well-argued case that it is wrong.

Specifically, the author would welcome:

• Evidence that the correlation tax is smaller than claimed — measurements of what cross-domain analytical work actually costs in organizations that have instrumented it.
• Demonstrations that incumbent architectures already close the gap — particularly from practitioners running mature data fabric or streaming platforms who find that cross-domain relationship discovery is not, in fact, a persistent problem.
• Counter-arguments to the governance model — especially from privacy engineers and regulators on whether purpose intersection and prohibited-inference registries are workable or merely plausible-sounding.
• Any partial implementation experience, successful or otherwise. Reports of failure would be more informative than reports of success, and the author commits to reflecting them in subsequent versions.

The framework’s falsification conditions are stated explicitly in Chapter 38 and §39.6 precisely so that this invitation can be taken up rigorously.

F.5 Acknowledgement of Standing on Others’ Work #

Every foundational idea in this framework belongs to someone else. The log as a unifying abstraction is Kreps’. Event-time semantics and watermarks come from the Dataflow lineage. Complex event processing predates this work by two decades. Federated computational governance is Dehghani’s. Knowledge graphs, retrieval augmentation, calibration methods, entity resolution theory, and the CAP and PACELC results all belong to the researchers cited in Chapter 42.

What this document attempts is composition and a shift of emphasis — placing relationships rather than records at the centre of the architecture. If that shift proves useful, the credit is distributed across the hundred and fifty works in the reference list. If it proves mistaken, the error is the author’s alone.

---

---

Return to: Cover · Table of Contents