0 Fork 0

Files

T

dennisthiessenandClaude Opus 5 ce5b7f511f fix(F07): propagate recorded corrections into active writing sources

The review's key point is that the error ran BOTH ways.

Overstatement, narrowed to canonical scope:
- BS-1 variants claimed "eliminated manual wafer inspection", "no manual
  intervention" and "unattended quality monitoring". Canonical BS-1 supports a
  qualitative reduction whose size is unverified. Reworded; the scope limit is
  now stated inline where the variants are written.
- bundle_data_platform called BS-1 a "zero-downtime constrained deployment".
  Integration into a fab that runs continuously is not a measured zero-downtime
  property.
- experience_swisscom carried "grounded retrieval" as a skill and ATS keyword
  while SW-8 records retrieval implementation as explicitly unknown, and its
  reframing note invited "evaluation discipline" although formal model
  evaluation is a globally forbidden pattern. Both removed.
- PyTorch sat in bundle_ml_ai_engineer as a bold tool and an ATS must-include
  with no certification context - the exact way the context gets lost.

Undervaluation, the half that quietly costs him roles:
- BS-6 (Linux administration, intensive Ansible incl. self-written extensions)
  has been canonical since 2026-08-27 but appeared in NO bundle, so the normal
  selection paths could not see it. Added to the data-platform and data-engineer
  bundles with its currency limit (employer-side ended Dec 2022; PP-3 is
  personal) and the standing GitOps prohibition.
- Linux, Ansible and Bash were missing from skills_taxonomy entirely.
- skills_taxonomy still said the IBM certification year was "not recorded" and
  that PyTorch needed verification; both were settled on 2026-08-27.

Not rewritten, deliberately: the NVIDIA and RUAG cover letters state the
production line could not be stopped. That exceeds the canonical record but is
plausible and both are already submitted. Recorded as an OPEN ITEM in
experience_bosch.md to be confirmed with the user before any reuse.

Validator PASS, 31/31 tests.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Qej5SBvhcbGXHwLqNWiskM

2026-09-11 19:37:36 +02:00

8.2 KiB

Raw Blame History

Bundle: ML / AI Engineer

Target employers: AI product companies, R&D teams Tier: 2 — strong with targeted emphasis Config key: bundle_ml_ai_engineer.md

S1: Role Profile & Priority Matrix

Positioning: Dennis is an ML Engineer who specializes in the infrastructure and deployment side of ML — not ML research or model training from scratch. His flagship signal is containerizing and orchestrating ML inference into a 24/7 semiconductor production environment (Bosch) — a production ML deployment in one of the most demanding operational contexts imaginable. This is paired with NLP/speech recognition research (Fraunhofer ARTUS), early deep learning exposure (IBM AI Engineering, Udacity AI for Trading), and current AWS/Kubernetes ownership at Swisscom.

Honest framing: Dennis is strongest in MLOps and ML infrastructure. He has real but limited model development experience (ARTUS NLP, image classification at Bosch). Do NOT position as a research ML scientist or model architect. DO position as "ML Engineer who can own the full deployment lifecycle and build reliable production ML infrastructure."

Key differentiator: Production ML in a 24/7 constrained environment (24h uptime, no deployment windows, hardware-in-the-loop) is a rare and credible signal.

Priority Matrix

Priority	Achievement IDs	Rationale
HIGH	BS-1, SW-3, SW-1, SW-2	Production ML deployment, K8s infrastructure, AWS data lake, pipeline ownership
MED	FC-2, SW-5, BS-2, BS-3, BS-4	NLP research, DevSecOps, data services, app ownership, observability
LOW	SW-4, FC-1, FC-3, VZ-1, VZ-2, GN-1, GN-2	Not core ML signal

2-page resume bullet allocation (typical):

Swisscom: 3 bullets (SW-3, SW-1, SW-2)
Bosch: 3 bullets (BS-1, BS-2 or BS-3, BS-4)
Fraunhofer: 1–2 bullets (FC-2 ARTUS; FC-1 if CI/CD is relevant)
Vizrt: 1 bullet (compressed; or omit)
Generali: 1 bullet (GN-1 or omit)

S2: Summary Guide

Headline pattern:

"ML Engineer | MLOps · Kubernetes · AWS | Production ML Deployment, Data Infrastructure & Applied NLP"

Building blocks:

"production ML deployment" or "ML inference at scale"
"containerized ML pipelines" (Docker, Kubernetes, Ansible)
"ML infrastructure and data platform" (AWS, Airflow, Kafka)
"applied NLP / speech recognition" (Fraunhofer ARTUS — hedged)
"AWS Certified Solutions Architect" (infra credibility)

Tone: Pragmatic ML engineer who ships ML to production. Operational rigor. Infrastructure-first mindset. Not a researcher — an engineer who makes ML run reliably.

Avoid:

Claiming deep model research or novel architecture contributions
Overstating NLP depth (ARTUS was contributing developer, not solo)
Leading with test automation or BI/analytics framing
"Passionate about AI" clichés

S3: Achievement Reframing Map

ID	Default Framing	This Role's Framing	Key Metric / Signal
BS-1	ML inference containerization	LEAD bullet — "Designed and deployed ML inference pipeline (Docker, K8s, Ansible) into 24/7 semiconductor fab; automated image-based defect classification"	Production ML in constrained 24/7 environment
SW-3	K8s + GitLab CI/CD	"Deployed and operated ML-ready Python applications on Kubernetes with GitLab CI/CD automation — production-grade containerized delivery"	K8s ownership = MLOps infrastructure
SW-1	AWS migration	"Migrated his domains' pipelines onto Swisscom's AWS platform (S3, Glue, Athena/Iceberg, Redshift, Airflow) — the data layer ML workloads depend on" — corrected 2026-08-26, previously read "Built cloud-native data infrastructure… the scalable data layer", pairing a full-ownership verb with an org-scale object (SW-1 `forbidden`) and asserting unverified scale	AWS data lake for ML workloads
SW-2	Component Owner	"Owned business-critical ETL pipelines (Oracle/Kafka → Teradata) — reliable data supply for downstream ML and analytics"	Data reliability for ML input
FC-2	ARTUS NLP	"Contributed ML and speech recognition components to ARTUS — Fraunhofer research project targeting automatic sea rescue transcription"	Applied NLP in safety-critical domain
BS-4	ELK PoC	"Delivered anomaly detection PoC: ELK + Kafka pipeline with Grafana/Prometheus monitoring — ML-adjacent signal processing"	Anomaly detection / observability
SW-5	Security Champion	Omit by default. If explicitly relevant: "2025/2026 team Security Champion"	Security exposure only; not responsible-AI, model-security or compliance ownership
BS-2	Data services	"Built data services (Python/Java/C#) over OracleDB and Hadoop enabling ML model input pipelines in semiconductor manufacturing"	Data infrastructure for ML

S4: Skills Guide

Bold tools (resume Technical Skills section): Python, Kubernetes, Docker, AWS (S3 · Glue · Athena · Airflow), Kafka, Scikit-learn (PyTorch and TensorFlow/Keras are certification-context-only in claims.json — list them only beside the IBM AI Engineering certificate, never as production tooling.)

Must-include skills (ATS match):

Python, MLOps, ML inference deployment
Docker, Kubernetes, Ansible
AWS (S3, Glue, Athena, Redshift, Airflow), Apache Iceberg
Apache Kafka
Scikit-learn, Pandas, NumPy
PyTorch / TensorFlow only with certification context — see the note above; a bare ATS listing is the exact way the context gets lost
CI/CD (GitLab, Jenkins)

Nice-to-have (include if JD mentions):

NLP, speech recognition (hedge — contributing developer at Fraunhofer)
TensorFlow/Keras (IBM cert — familiarity level; note if JD requires)
PySpark (Swisscom — Spark ML adjacent)
Grafana, Prometheus (monitoring ML systems)
Apache Spark ML (familiarity via cert)

Omit:

BDD, Selenium, HP Quality Center (testing — irrelevant)
RPA/UIPath, Camunda (process automation — irrelevant)
Tibco Spotfire (BI tool — irrelevant)
SAP BODS (legacy — omit for ML audience)

Certifications to highlight:

AWS Certified Solutions Architect – Associate → HIGH (infra credibility for MLOps)
AI for Trading Nanodegree (Udacity) → MED (quantitative ML exposure)
IBM AI Engineering Specialization → MED (TensorFlow, Keras, Spark ML depth signal)
Data Engineering with AWS → supporting

S5: Cover Letter Guide

Institution type: AI product company, applied ML team, or R&D engineering team (not pure research lab)

Opening hook pattern:

"My most defining ML engineering project was at Bosch Semiconductor in Dresden: I designed and deployed the strategy for containerizing ML inference models into a 24/7 wafer production line — orchestrating with Kubernetes and Ansible, running Docker-containerized image classification models continuously against production data, with no downtime tolerance. That production constraint is what sharpens MLOps thinking. [Tie to their ML deployment challenge]."

Key narrative thread:

Production ML ownership — BS-1: not a research demo, not a notebook — ML running uninterrupted in a semiconductor fab. Establish this credibility early.
Infrastructure that ML depends on — SW-1 + SW-2: the AWS data lake and reliable Kafka/ETL pipelines that feed ML models. Shows full-stack ML engineering thinking.
Applied ML research background — FC-2: Fraunhofer ARTUS NLP project establishes early research exposure (hedge: contributing developer)
Certs as credibility — AWS SAA + IBM AI Engineering + Udacity AI for Trading: show structured ML and cloud learning

"Why them" angle to research:

What ML problems are they solving? Map to defect detection (vision ML), NLP (Fraunhofer), or infrastructure (Swisscom)
How mature is their MLOps? If early-stage → your Bosch experience building from scratch is directly relevant
Are they AWS-heavy? → Your migration and cert are strong signals