BASSAM®

AI Infrastructure Engineer, Prague

BassamMohamed

I build AI systems that take repetitive work off people's plates.

Portrait of Bassam Mohamed

With a sharp eye for repetition, Bassam turns manual work into a system.

(About Bassam) He works inside real teams, close enough to see where repetition, handoffs, and fragile workarounds have become normal. Bassam maps the process, finds the leverage point, and builds the smallest useful system that removes the work instead of moving it elsewhere.

His work can take the form of an AI agent, retrieval layer, data pipeline, browser tool, or full product. He carries it from problem framing and architecture through interface, production code, deployment, and the operating model people need to trust it.

(work)

# 01
  • React
  • Electron
  • macOS
  • Local-first
s8vr

Send the invoice. Get back to work.

Local app · macOS

Visit website

A downloadable Mac app for creating and managing invoices locally, without a SaaS subscription.

Project summary

s8vr.app

A downloadable Mac app for creating and managing invoices locally, without a SaaS subscription.

# 02
  • Swift
  • Python
  • Agent memory
  • Open source
.scrollback

Close the tabs. Keep the work.

Open source · macOS

View GitHub

A macOS app that turns scattered agent chats into work summaries you can review, save, and carry forward.

Project summary

Scrollback

A macOS app that turns scattered agent chats into work summaries you can review, save, and carry forward.

# 03
  • Agents
  • Product
  • Experimentation
  • UX review
Experiment Proposer

Past results. Next experiment.

Internal project · Conceptual view

A weekly agent that turns prior results into ranked, approval-ready experiment briefs while separating evidence from estimated impact.

Project summary

AI Experiment Proposer

A weekly agent that turns prior results into ranked, approval-ready experiment briefs while separating evidence from estimated impact.

My work brought multiple sources and reusable AI skills into experiment planning. The workflow separates prior evidence from estimated impact, with human review before approval. This is a public-safe summary of an internal project; no measured uplift is claimed.

# 04
  • OpenMetadata
  • Data governance
  • AI workflows
AI-ready Data Catalog

Clear definitions. Shared understanding.

OpenMetadata · Conceptual view

I helped standardize metric definitions in OpenMetadata so people and AI agents could work from consistent, owned definitions.

Project summary

AI-ready Data Catalog

I helped standardize metric definitions in OpenMetadata so people and AI agents could work from consistent, owned definitions.

My contribution included registering metrics, auditing conflicting definitions, coordinating fixes with owners, and identifying gaps in the AI-ready standard and its supporting Claude workflow. This was collaborative catalog-improvement work. Some remediation remained open; this is not a claim that I rebuilt the whole platform.

# 05
  • Python
  • Streamlit
  • BigQuery
  • Keboola
Metric Watchdog

Spot the change. Find the context.

Internal project · Conceptual view

An internal monitoring app that adds context to critical metrics and makes morning triage a quick visual review.

Project summary

Metric Watchdog

An internal monitoring app that adds context to critical metrics and makes morning triage a quick visual review.

The project combines metric monitoring with context for morning review and daily announcements. Its stack includes Python, Streamlit, BigQuery, and Keboola. Internal data and identifiers are omitted here.

(loadout)

01

AI Infrastructure

Claude systems, RAG pipelines, autonomous agents

Claude API / Pinecone / n8n
02

Data & Analytics

Petabyte scale pipelines that flag their own anomalies

SQL / BigQuery / dbt
03

Engineering

Full stack product work, built to survive production

TypeScript / React / Python
04

Product & Strategy

From ambiguous problem to dated, shipped roadmap

Agile / HCD / Figma
05

DevOps & Infra

Containerized, monitored, deployed on real hardware

Docker / Linux / Hetzner
06

Security & Ops

Zero trust architecture from the first conversation

OWASP / RBAC / OAuth2
01 / 03

(signals)

“Bassam is the kind of person who sees you doing something manually and just fixes it. Not because anyone asked. Because watching waste bothered him.”

DarylMSF

“Bassam changed how we shared our work. He made the results visible, turned lessons into reusable practice, and helped other teams build on what we learned.”

PavelGroupon

“Bassam is an automation builder who turns ambiguity into clear systems. He finds the right questions, connects the moving parts, and ships tools the team can actually use.”

KaramUber

(reach out)

Let’s talk.

If you’re building with AI, automating work, or shipping a useful product, I’d like to hear about it.

hello@bassam.czConnect on LinkedIn