dennisthiessen
dd2f0308c5
Observability-spine framing: Bosch telemetry/observability PoC promoted to lead, Swisscom Iceberg lakehouse + on-call/RCA. Real Ashby JD (verbatim). Tier 1+2 critique fixes applied; Vizrt low-latency skipped per user. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
18 KiB
18 KiB
Session: Snowflake — Senior Software Engineer, Enterprise (Observe by Snowflake)
Status: Phase 0: DONE — awaiting user confirmation before Phase 1
Created: 2026-06-06
JD source: output/Snowflake_Observe_Enterprise/JD_snowflake_observe_enterprise.txt (LIVE PULL 2026-06-06 via Ashby public posting API, board slug snowflake, job id 3eea87fa-... — authoritative, verbatim. NOT a reconstruction.)
Output folder: output/Snowflake_Observe_Enterprise/
JD Info
- Role: Senior Software Engineer, Enterprise — Observe by Snowflake
- Company: Snowflake (Observe team — acquired ~$1B, announced Jan 8 2026, closed)
- Location: CH-Zurich (Zürich, Switzerland) → Bern-commutable hybrid (clears comp bar)
- Bundle (primary): Staff / Senior Data Engineer (
bundle_data_engineer.md) — Tier 1, strongest evidence - Bundle (secondary): Data Platform / Infra (
bundle_data_platform.md) — for the observability/backend-platform framing - Format: Resume (2-page, resume.cls) + 1-page cover letter
- Comp (LISTED ON POSTING): CHF 176K–253K base — clears the 180k all-in bar on base alone (+ Snowflake equity/bonus). Verified from the JD.
- Why this one: best-fit role in the whole 2026-06 search — backend/platform SWE on an observability lakehouse; Python/Java explicitly accepted (NO C++ gate, unlike the dropped Google reqs); bonus points read like Dennis's résumé (observability, Iceberg, AI-assisted workflows, SRE/RCA).
JD Analysis
Requirements
| # | Requirement | Match | Evidence |
|---|---|---|---|
| 1 | 5+ yrs professional SWE, backend systems / distributed platforms | DIRECT | 11+ yrs; backend services, data pipelines, distributed processing |
| 2 | Designing & shipping production systems at scale (high throughput, low latency, enterprise reliability) | DIRECT (scale caveat) | Telco ETL at high volume + on-call SLA; Vizrt real-time low-latency backend. NOT petabyte — frame honestly |
| 3 | Strong proficiency in a server-side language: Go, Java, Scala, Python, or similar | DIRECT | Python expert, Java strong (Go = learning; Scala n/a — Python/Java satisfy "one or more") |
| 4 | Work independently; own complex problems scoping → delivery | DIRECT | Component Owner + Application Owner; end-to-end ownership under on-call |
| 5 | Collaborate across engineering, product, infrastructure (fast-moving, high-ownership) | DIRECT | Cross-functional fab ops, B2B stakeholders, agile DevOps |
| 6 | Strong analytical / structured problem-solving; ships customer-facing work | DIRECT | RCA under on-call; data products for downstream consumers |
| 7 | BS/MS CS or equivalent | DIRECT | M.Eng. Computer Aided Engineering (Software Design & Eng.) |
| 8 | BONUS: observability / telemetry pipelines / tracing+logging (OpenTelemetry, Prometheus, Jaeger) | DIRECT | Bosch observability stack: ELK + Kafka + Grafana + Prometheus + Loki — designed it for anomaly detection + reliability |
| 9 | BONUS: Apache Iceberg, Parquet, open data lakehouse | DIRECT | Swisscom Iceberg / Athena / S3 lakehouse + Data Mesh data products |
| 10 | BONUS: AI-assisted dev tools / LLM-integrated workflows | DIRECT/BRIDGE | Kiro, Copilot, LiteLLM (LLM gateway), custom GPTs (verified toolchain — NOT LangChain) |
| 11 | BONUS: SRE practices, incident response automation, RCA tooling | DIRECT | On-call SLA ownership + 3rd-level root-cause analysis; observability for reliability |
| 12 | BONUS: open-source contributions to observability / data infra | GAP (minor) | No observability OSS. (Writes Solidity in free time — not relevant here.) Bonus-only, low weight |
ATS Keywords
- Core: software engineer, backend, distributed systems, distributed platforms, production systems at scale, APIs, high throughput, low latency, reliability, end-to-end ownership
- Observability: observability, telemetry, logs/metrics/traces, OpenTelemetry, Prometheus, monitoring, incident response, root-cause analysis, SRE
- Data/lakehouse: Apache Iceberg, Parquet, data lakehouse, data pipelines, ingestion, query, columnar storage
- Languages: Python, Java (Go learning), SQL, Scala (n/a)
- AI: AI-assisted, LLM-integrated workflows, AI SRE, agentic
- Soft: cross-functional, code review, mentorship, architecture, customer-focused
Gap Assessment
- Direct: SWE years, backend/distributed, Python/Java, ownership, collaboration, problem-solving, education, observability/telemetry, Iceberg/lakehouse, AI-assisted workflows, SRE/RCA.
- Bridge: AI-assisted workflows (real toolchain, but lighter than core eng); low-latency (Vizrt real-time is the proof point — telco is throughput-heavy).
- Gap / honest caveats: (a) not petabyte/Observe-scale telemetry — don't overclaim, keep honest qualitative scale; (b) no observability OSS (bonus only); (c) "Senior" vs his Staff title — non-issue given 176–253k band; (d) no fabricated metrics (user directive — facts only).
Company Context
- Mission: Snowflake = the data/AI company ("AI Data Cloud"). Observe by Snowflake = AI-powered observability platform Snowflake acquired (~$1B, Jan 2026, its largest deal), expanding into the $50B+ IT-operations-management market.
- Architecture: AI SRE + Observability Context Graph + Telemetry Lakehouse Foundation on Apache Iceberg + OpenTelemetry; ingests logs/metrics/traces/events at petabyte scale with compute-storage separation, "detection → root cause 10x faster." Customers incl. Capital One, Topgolf, Dialpad.
- This role (Enterprise team): build/own core features, APIs, business logic; production-grade backends ingesting massive telemetry; evolve the open Iceberg lakehouse model; partner with SREs on AI-assisted RCA; code review + mentorship.
- Culture: "AI-native," low-ego, high-ownership, startup velocity inside a big data platform; ships fast, customer-focused.
- "Why them" angle: Dennis has lived both sides of what Observe sells — he built a production observability stack (Bosch ELK/Kafka/Grafana/Prometheus/Loki) AND owns telemetry-scale data pipelines on an Iceberg lakehouse end-to-end with on-call/RCA. That operator-who's-been-the-AI-SRE's-customer profile is rare. Zürich is local-commutable.
Framing Strategy
- Lead narrative: Backend/platform engineer who builds and owns production data + telemetry pipelines end-to-end — observability stack, Iceberg lakehouse, APIs/services on Kubernetes — with on-call SLA / root-cause ownership and a Python/Java core.
- KEY FLIP vs the shelved Google/Isovalent packages: here observability is the SPINE, promoted to the lead (it was demoted there). Lead Bosch with the observability/telemetry bullet; lead Swisscom with ownership + Iceberg lakehouse.
- Reframing map:
- Bosch ELK/Kafka/Grafana/Prometheus/Loki stack → "telemetry pipeline + observability platform: ingest logs/metrics, anomaly detection, reliability" (promoted lead)
- Swisscom Iceberg/Athena/S3 + Data Mesh → "open data lakehouse (Apache Iceberg) + governed data products"
- Python services on K8s + APIs → "production-grade backend services and APIs"
- On-call SLA + 3rd-level RCA → "SRE practices, incident response, root-cause analysis"
- Kiro / Copilot / LiteLLM / custom GPTs → "AI-assisted / LLM-integrated workflows" (the AI-SRE angle)
- Kafka / PySpark → "high-throughput distributed processing at telemetry scale"
- Vizrt real-time transcoding → "low-latency real-time backend" (the low-latency proof point)
- Emphasize: observability/telemetry, Iceberg/lakehouse, Python/Java backend + APIs, end-to-end ownership + on-call/SRE/RCA, distributed processing, AI-assisted workflows.
- Downplay: pure DevOps/IaC (unless supporting), C++ (NOT needed here — Python/Java are the ask; don't oversell per
feedback_cpp_emphasis), BDD/RPA early career, semiconductor-domain specifics (keep transferable). - CL hooks: Snowflake's $1B Observe acquisition (Jan 2026); AI-powered observability on Iceberg + OpenTelemetry; Dennis built an observability stack AND owns Iceberg-lakehouse pipelines with on-call; the AI-SRE/agentic angle; Zürich-local.
- User directives / caveats: no fabricated tools (Kiro/Copilot/LiteLLM/custom GPTs only — NOT LangChain); no fabricated metrics (facts only — user reaffirmed 2026-06); don't overclaim petabyte scale; honest low-latency framing; Go = "learning"; Python/Java are the strengths.
Critique Context
- Reviewer persona: Observe/Snowflake senior backend engineer (often ex-startup, observability-domain) — values distributed-systems depth, telemetry/observability fluency (OpenTelemetry, Prometheus), end-to-end ownership, pragmatic shipping, Iceberg/lakehouse knowledge. Skeptical of buzzwords; wants evidence of building and operating real backend systems at scale.
- Competitive landscape: backend/distributed-systems engineers, observability-product engineers (ex-Datadog/Grafana/Elastic), data-platform engineers. Dennis's edge: the rare operator who built an observability stack AND owns telemetry-scale pipelines on an Iceberg lakehouse end-to-end with on-call/RCA — he's lived data-infra + observability + SRE at once. His risk: not petabyte-scale, no observability OSS, "Senior" vs Staff (minor).
- Domain vocabulary: telemetry (logs/metrics/traces/events), OpenTelemetry, Prometheus, ingestion, query execution, columnar storage, Apache Iceberg, Parquet, lakehouse, compute-storage separation, SLO/SLA, incident response, root-cause analysis, throughput/latency, distributed systems.
Cover Letter Plan
- Institution type: Industry (big-tech / data-infra, recently-acquired startup team)
- Paragraph count: 3 paragraphs, ~250–300 words (1 page)
- P1 hook: Snowflake's $1B Observe acquisition + AI-powered observability on Iceberg/OpenTelemetry; Dennis's combo of building an observability stack and owning telemetry-scale pipelines.
- P2 evidence: Bosch observability stack (ELK/Kafka/Grafana/Prometheus/Loki) + Swisscom Iceberg/Athena lakehouse + Python/Java backend services & APIs + on-call SLA / 3rd-level RCA (SRE).
- P3: AI-assisted workflows (LiteLLM/custom GPTs — the AI-SRE angle) + honest scale framing + Zürich-local; active close.
- Jargon level: Technical.
- "Why them" hook: Dennis has lived the exact intersection Observe sits at — data infra + observability + SRE — and ships end-to-end.
Output Files
- Resume:
output/Snowflake_Observe_Enterprise/e2e_snowflake_observe_enterprise_resume.tex - Cover Letter:
output/Snowflake_Observe_Enterprise/e2e_snowflake_observe_enterprise_cover_letter.tex - Critique:
output/Snowflake_Observe_Enterprise/critique_snowflake_observe_enterprise.md
Bullet Plan (Phase 1 — PROPOSED 2026-06-06)
Position title themes (observability/backend-platform flip):
- Swisscom: Backend Data Platform, Lakehouse & Pipeline Reliability
- Bosch: Observability, Telemetry Pipelines & Production Data Services (observability PROMOTED to lead)
- Fraunhofer: Distributed Services & Applied ML Engineering
- Vizrt: Distributed Real-Time Backend at Broadcast Scale
- Generali: Build Automation, CI/CD & Java Backend
Swisscom (6) — Staff Data, Analytics & AI Engineer
| ★/o | ID | Achievement | JD Match |
|---|---|---|---|
| ★ | SW-2 | Component Owner — Fulfillment/Product-Analysis ETL (Oracle/Kafka→Teradata, Python); on-call SLA + governance | Direct (ownership + SRE) |
| ★ | SW-1 | AWS migration → cloud-native lakehouse w/ Apache Iceberg (S3/Glue/Athena/Iceberg/Redshift/Airflow) | Direct (Iceberg/lakehouse — bonus #9) |
| ★ | SW-7 | Data Mesh + metadata/governance — foundation for AI/agentic workflows | Direct (lakehouse gov + AI-assisted #10) |
| ★ | SW-3 | Python services on Kubernetes + GitLab CI/CD — containerized backend delivery | Direct (backend services/APIs) |
| ★ | SW-6 | PySpark distributed processing over the Data Lake (high-throughput) | Direct (scale/throughput) |
| ★ | SW-4 | B2B data products + automation + 3rd-level RCA under on-call | Direct (RCA/reliability #11) |
| x | SW-5 | Security Champion | Skip (low relevance; correction-limited to 2025/26, not an award) |
Bosch (4) — (Senior) Data Engineer — observability PROMOTED to lead
| ★/o | ID | Achievement | JD Match |
|---|---|---|---|
| ★ | BS-4 | Observability / anomaly-detection PoC — ELK + Kafka + Grafana/Prometheus/Loki (HONEST PoC framing) | Direct (observability/telemetry — bonus #8, the SPINE) — LEAD |
| ★ | BS-2 | Data services in Python/Java (+C#) over Oracle + Hadoop/Impala; query optimization (columnar) | Direct (backend + Python/Java + query) |
| ★ | BS-1 | Containerized ML inference (Docker/K8s/Ansible) into 24/7 fab, no downtime | Bridge (production reliability + AI) |
| ★ | BS-3 | Application Owner — SLOs, cross-functional, training/docs | Direct (ownership/SLO — SRE) |
Fraunhofer (3) — Research Software Engineer
| ★/o | ID | Achievement | JD Match |
|---|---|---|---|
| ★ | FC-3 | Microservices + REST APIs (Express.js, Docker) — MISSION data-exchange platform | Med (backend/APIs) |
| ★ | FC-1 | SCEDAS + independently set up Jenkins CI/CD + build automation | Med (CI/CD initiative) |
| o | FC-2 | ARTUS ML/NLP (hedged "Contributed") | Med (ML/AI breadth) |
Vizrt (2) — DevOps Engineer
| ★/o | ID | Achievement | JD Match |
|---|---|---|---|
| ★ | VZ-1 | Distributed real-time transcoding backend (Python, legacy C++) — CNN/BBC/Al Jazeera | Direct (low-latency/distributed — the low-latency proof point) |
| ★ | VZ-2 | A/V test suite + CI/CD quality gates | Bridge (quality/reliability) |
Generali (2–3) — IT Consultant (Hamburg)
| ★/o | ID | Achievement | JD Match |
|---|---|---|---|
| ★ | GN-1 | Introduced BDD test automation + technical ownership; trained Java Community | Weak (initiative) |
| ★ | GN-3 | Java/J2EE features (PIA-Postkorb), XLDeploy, Apache Camel/Spring Boot PoC | Weak-Med (Java backend) |
| o | GN-2 | UIPath RPA PoC | Page-fill only |
Recommended total: 17 (SW ×6, BS ×4, FC ×3, VZ ×2, GN ×2 [GN-1+GN-3]). Budget target ~18–20 → GN-2 (RPA) is the optional 18th for page-2 fill, decided at the Phase 2 page-fill gate. Reuse note: same underlying bullets as the shelved Isovalent/Google packages, but TEXT retuned to observability/telemetry/Iceberg/SRE vocab and REORDERED — observability promoted (Bosch lead), C++ de-emphasized (Python/Java are the ask).
Status
- Phase 0: DONE (real JD pulled verbatim via Ashby API; research done)
- Phase 1: DONE (18 bullets confirmed: SW-2/1/7/3/6/4, BS-4/2/1/3, FC-3/1/2, VZ-1/2, GN-1/3/2)
- Phase 2 Resume: DONE — compiled clean 2 pages (resume.cls copied in); no OVER violations; page 2 ~60% full (not padded). Observability promoted to Bosch lead; BS-4 honest "proof of concept"; SW-7/SW-1 scoped per Scope Discipline; no fabricated metrics; no OpenTelemetry claim (not in his toolset).
- Cover Letter: DONE — compiled clean, 1 page, ~294 words (3 paragraphs, industry). Hooks web-verified (Observe ~$1B acquisition Jan 2026; AI SRE + Context Graph + Telemetry Lakehouse on Iceberg/OpenTelemetry). 0 em-dashes; AI-fingerprint scan clean; package cohesion verified (all claims traceable to resume bullets); no fabricated metrics, no LangChain, no OpenTelemetry self-claim, scope-disciplined.
- Critique: CURRENT — Pass 1 84.0/100 → Pass 2 86.0/100 (edits applied 2026-06-06). Strong, submit-ready. No fatal gaps (all 7 hard quals DIRECT). CL unchanged (no edits needed).
- Applied: Tier 1 (B2 -ing ending → concrete; B3 grammar "that" + "Modeled" reword) + Tier 2 (SRE practices + Parquet to skills; code review into B10; "Built" openers 6→4). Recompiled clean, 2pp, no orphans, all bullets ≤218.
- Skipped per user: Vizrt low-latency (not significant, long ago). This leaves ~1pt on the table; ceiling ~88 (structural: not petabyte-scale, no observability OSS).
- SUBMITTED 2026-06-06 (~86/100). Submission PDFs:
Dennis_Thiessen_Resume.pdf,Dennis_Thiessen_Cover_Letter.pdf. Applied at: https://jobs.ashbyhq.com/snowflake/3eea87fa-73c8-46dc-b69b-7beb438b48d8 - "currently learning Go" confirmed OK to keep (user, 2026-06-06).
Output Files (added)
- Resume:
output/Snowflake_Observe_Enterprise/e2e_snowflake_observe_enterprise_resume.tex(+ .pdf, 2pp) - Cover Letter:
output/Snowflake_Observe_Enterprise/e2e_snowflake_observe_enterprise_cover_letter.tex(+ .pdf, 1pp)
CL Hook Verification (web-verified 2026-06-06)
- Observe acquisition ~$1B, announced Jan 8 2026, Snowflake's largest deal (surpassing $800M Streamlit), closed Feb 2 2026 → TechCrunch / DevOps.com / Snowflake press release.
- Architecture: AI SRE + Observability Context Graph + Telemetry Lakehouse Foundation on Apache Iceberg + OpenTelemetry; observability data written directly to Iceberg tables; "root cause up to 10x faster" → Snowflake press release. (CL references Iceberg lakehouse + AI SRE; does NOT claim OpenTelemetry as Dennis's tool.)
Generation notes / accuracy flags
- JD is REAL (verbatim Ashby pull) — no reconstruction. Comp CHF 176–253k base confirmed on posting.
- Observability is the SPINE here (promoted), opposite of the Google/Isovalent framing — reuse bullets but re-tune to observability/telemetry/Iceberg vocab and reorder.
- No fabricated tools (Kiro/Copilot/LiteLLM/custom GPTs; NOT LangChain). No fabricated metrics (facts only). Education dates KB-corrected. Generali = Hamburg, Bosch = Dresden.
- "currently learning Go" — verify still true before submitting (shared across packages).