Software Engineer, AI & Backend @ Tend

Cory Janowski

11+ years building backend systems that have to keep working, six of them at Apple. These days: agentic LLM workflows

Fort Wayne, Indiana

Email me LinkedIn GitHub
Cory Janowski

// about

Systems that have to keep working

Backend and AI software engineer, including six years at Apple. I build the systems that have to keep working: data pipelines, APIs, agentic LLM workflows, and the observability and on-call practice around them.

At Tend I design GraphQL and AWS backend features and production agentic task loops in Python and TypeScript. Tuning SQS and API retry, backoff, and jitter logic cut timeout errors and failed jobs by over 90%, and fixing bottlenecks in our multi-agent architecture reduced outages 25%.

fewer timeouts & failed jobs at Tend
90%
compressed payloads through Go ingestion
50 GB+
data sources integrated at Abre
150+
profile.ts
1const cory = {
2  role: "Software Engineer, AI & Backend",
3  at: "Tend",
4  yearsInIndustry: 11,
5  topSkills: [
6    "Distributed Systems",
7    "Python",
8    "Go",
9    "TypeScript",
10    "AWS",
11  ],
12  ships: [
13    "agentic LLM workflows",
14    "distributed systems",
15    "data pipelines",
16    "observability & on-call",
17  ],
18  openSource: [
19    "github.com/cjanowski/claude-activity-graph",
20  ],
21} satisfies Engineer

// experience

Where I've worked

  1. Software Engineer · Tend

    May 2026 – Present · 7 mos

    Nashville, Tennessee

    • Design and build backend features and GraphQL APIs on AWS for a TypeScript and Python full-stack platform.
    • Architect production agentic task loops and multi-agent systems with LLM libraries, automating workflows that previously required manual intervention.
    • Tuned SQS and API retry, backoff, and jitter logic, cutting timeout errors and failed jobs by over 90%.
    • Led debugging and optimization of the distributed agent architecture, reducing system outages 25%.
    • Built AI verification and status labeling for the insurance-tracking pipeline, improving its reliability.
    • Built admin user backends with Cognito integration and metrics dashboards for observability and proactive monitoring.
    • Led integration of Claude into the codebase and backend (skills, context, tooling).

  2. Software Engineer II · Abre

    Jan 2024 – Apr 2026 · 2 yrs 4 mos

    Cincinnati, Ohio

    • Architected fault-tolerant data-ingestion infrastructure in Go processing 50 GB+ compressed payloads, with custom binary codec parsers and exponential-backoff retries; the design patterns were later adopted by three engineering teams.
    • Designed service architecture for a cloud data integration platform connecting 150+ data sources into BigQuery via REST and gRPC APIs, reducing primary datastore load and enabling analytics queries.
    • Created scripted deployment templates that cut deployment time from hours to minutes.
    • Led the end-to-end update of the Data Monitoring Center, including drag-and-drop SFTP; support escalations for SFTP credentials fell 90%.
    • Mentored two new engineers and ran design and code reviews for the Data Integrations UI.
    • Managed the PagerDuty on-call rotation for P0 incidents, improved logging and alerting to reduce MTTR, and wrote three standardized incident runbooks.

  3. Engineer II, Reliability/QA · Total Expert

    Jun 2021 – Dec 2023 · 2 yrs 7 mos

    Minneapolis–St. Paul

    • Designed and implemented the team's observability strategy on DataDog and AWS: real-time dashboards, alerting, telemetry, and centralized logging across distributed microservices, establishing early issue detection.
    • Led a database performance initiative, re-architecting core SQL queries and cutting execution times 50% for high-frequency queries.
    • Built automation testing frameworks and monitoring for the platform.
    • Provided on-call site support, trained new hires, and led proactive bug hunting.

  4. Software Engineer, Cellular/Wireless · Apple

    Jun 2015 – Jun 2021 · 6 yrs 1 mo

    Cupertino, California

    • Engineered and automated a Python error-reporting pipeline for wireless testing workflows, reducing issue acknowledgment time to under 4 hours and enabling rapid triage of critical bugs.
    • Maintained backend cron and Apache Airflow systems for wireless network analytics at 99.9% job reliability.
    • Monitored and enhanced cellular baseband observability across test fleets.
    • Built Looker dashboards that turned complex wireless metrics into actionable insights for cross-functional stakeholders, improving data accessibility 40%.

// projects

Things I've built on the side

featured

claude-activity-graph

A coding agent that keeps its own engineering record as a typed knowledge graph. A rule, a skill and a Stop hook make Claude Code log tasks, decisions and lessons into Obsidian; stdlib tools lint the graph and measure how lessons get reused across projects.

PythonClaude CodeObsidiand3
SourceLive graph

// skills

Stack

Languages

  • Python
  • Go
  • TypeScript
  • Java
  • SQL

Cloud & infra

  • AWS
  • GCP
  • Docker
  • Kubernetes
  • Terraform

Data

  • Postgres
  • BigQuery
  • Airflow
  • GraphQL
  • gRPC

Observability

  • DataDog
  • Grafana
  • OpenTelemetry
  • PagerDuty

AI & agents

  • LangChain
  • LangSmith
  • Claude Code
  • LLM agents
  • RAG

Highlighted: top skills on LinkedIn. Also: distributed systems design, on-call and incident response.

// education

Education & certifications

Woz U

University Diploma, Data Science

Jan 2020 – Dec 2020

Data science and machine learning program: NLP, TensorFlow, PyTorch, scikit-learn, Python, R, SQL.

Regis University

Coursework (non-degree), Computer Science

Jan 2017 – Jun 2018

Java I and II, programming and CS foundations.

  • Training Neural Networks in PythonMar 2023
  • Advance Your Skills in Deep Learning and Neural NetworksNov 2022
  • Advance Your Skills in AI and Machine LearningNov 2022