Bilal · AI & Digitalization Consultant who Implements · Germany

I build (agentic AI) systems
that deliver measurable outcomes.

Applied AI consulting and custom agentic tooling for engineering and management teams. From architecture and PoC through to production deployment.

Industry Context · The Production Gap

Most AI work never reaches production.

95% of enterprises report zero measurable profit impact from AI adoption. The shortfall is not in the models. It is in the engineering, the operations, and the discipline required to make a model survive contact with a real organisation.

I work in the gap between a working notebook and a system someone else can run on a Monday morning without calling me. That gap is where 80–85% of projects quietly die, and where the real economics live.

Now · Q1 2026 Onward

Operational analytics & workflow automation in Germany.

Phase II of an advisory and implementation engagement for a German small-medium enterprise (Mittelstand). Architected custom Node.js middleware bridging Toggl, Absence.io, and internal CRM tool APIs into a centralized Power BI analytics engine for executive leadership.

Power BI · Node.js · Express · REST APIs · Toggl · Absence.io · Render

CAPABILITIES

Four practices, one outcome: AI built for real-world impact.

/01

Agentic Systems

Multi-step LLM pipelines with tool use, retrieval, and evaluation. Built for production latency and cost constraints.

Stack:

Claude · Pydantic · DSPy

/02

BI & Analytics

KPI design, Power BI dashboards, Node middleware to bridge systems that don't speak to each other.

Stack:

Power BI · Node.js · DAX · Express · Render

/03

Research Engineering

Data infrastructure for academic and regulatory research: PDF ingestion, LLM extraction, reproducible pipelines.

Stack:

Python · Pandas · DuckDB · Claude

/04

Ops Automation

Internal tools that collapse hours of operations work: note-to-ticket, report generation, more.

Stack:

Python · Claude API · Notion API · Atlassian · Scriptable

Selected Work · Proven Outcomes

Systems engineered & shipped.

Selected case studies with verified, quantifiable results. Full index available at /projects.

  1. 01Advisory Engagement & Implementation · SME (Mittelstand)

    Unified Operations Middleware & BI Architecture

    A German small-medium enterprise (Mittelstand) with no end-to-end view of efficiency. Mapped their operational processes and engineered automated ETL pipelines unifying fragmented time-tracking (Toggl), HR management (Absence.io), and internal CRM tool into 8 executive Power BI dashboards with 23 DAX metrics.

    Read full case study
  2. 02Workflow Automation & Integration Systems

    Self-Hosted n8n Workflow Automation Infrastructure

    Production self-hosted n8n workflow automation platform orchestrating multi-service webhooks, scheduled cron triggers, and custom API integrations across local containerized nodes, enabling un-capped 24/7 background execution without commercial task quotas.

    Read full case study
  3. 03Research Infrastructure · TRR 266 (TUM × LMU × Bocconi)

    STOXX Europe 600 Climate Disclosure Pipeline

    A TRR 266 project needed financial statements and audit reports from six years of STOXX Europe 600 annual reports, machine-readable. Built an agentic data pipeline: automated fetching, PDF parsing, LLM-backed section extraction, and structured output into an NLP dataset with page-level provenance. Formed the empirical basis for SSRN 4763140.

    Read full case study
Operating Model · How I Work

Direct access. Scoped deliverables. Measurable outcomes.

  1. i.

    Direct collaboration.

    I work directly with decision makers. The shortest line between a problem and a working system.

  2. ii.

    Defined scopes.

    Engagements run on transparent, fixed-scope milestones (6 to 14 weeks). Clear technical specifications and deliverables agreed upon before implementation.

  3. iii.

    Operational-grade outputs.

    Success is measured by outcome: a live analytics engine, a verified agentic pipeline in daily operation, a system someone else can run without me.

  4. iv.

    Upfront feasibility.

    If a problem is ill-suited for current AI architectures or outside my core expertise, I identify that during the initial technical discovery call.

MANIFESTO

Engineering Axioms

  1. 01.DATA > OPINIONS.
  2. 02.AGENTS ARE EMPLOYEES' ASSISTANTS, NOT EMPLOYEES.
  3. 03.IF IT ISN'T MEASURED, IT ISN'T DONE.
  4. 04.SHIP, THEN POLISH. NEVER THE OTHER WAY ROUND.
  5. 05.SIMPLE PIPELINES BEAT CLEVER PROMPTS.
mbilal.works · personal hub at bilalm.me · Impressum · PrivacyBuilt to a DESIGN.md spec · Bricolage Grotesque + JetBrains Mono · MMXXVI