Data & AI Architect — fifteen years building data platforms where the numbers have to be right.

I work at the intersection of structured knowledge, modern AI retrieval, and data infrastructure.

• Knowledge retrieval systems — combining RDF triplestores, probabilistic logic (ProbLog), and vector search (pgvector) to build retrieval that handles both explicit relationships and semantic similarity.
• Macro-financial data infrastructure — collecting and analyzing inflation expectations, central bank credibility metrics, and macro indicators across 45+ emerging market countries. Live dashboard →
• KoyraCloud — open-source self-hosted PaaS: push-to-deploy on your own Docker Swarm. The deployment layer behind my products, including my consulting site.
• Data platform tooling — Dagster pipelines, DuckDB/Polars for local-first analytics, ClickHouse for time-series, all running on a 6-machine homelab (Proxmox + Docker Swarm).

Python (15+ yrs) · Dagster · Polars · DuckDB · ClickHouse · FastAPI · PostgreSQL · pgvector · RDF/SPARQL · ProbLog · Docker · Proxmox · Rust (learning)

• Local-First Analytics — my book on running analytics at warehouse speed without renting a cloud warehouse.
• From Stranger to Sold — my other book, for people who can build but find selling opaque. It maps the nine stages a stranger walks before buying, on the argument that upstream problems masquerade as downstream ones: weak positioning looks like a lead problem. Free, no signup — a 5-minute test that scores all nine and names your weakest, the first chapter, and field notes diagnosing one symptom each.
• Engineering lessons and observations on LinkedIn and X.

I take on a small number of engagements through Hikmah Technologies — production AI on your own data:

• Fractional data engineer — pipelines, a warehouse, dashboards your team can trust, part-time.
• AI automation engineer — approval-gated LLM workflows over your data (AEGIS is the open-source working example).

Free, no signup: the data platform teardown and the AI workflow teardown — the questions I ask before an audit or an LLM workflow goes to production. How engagements run — hours, payment, contracts — is written down at how I work.

Consulting inquiries: arshad@hikmahtechnologies.com