# Industry

> Software engineering across enterprise AI, energy, analytics, pharma, and open-source projects, with a career-wide overview of technologies and tools.

Source: [https://www.indrapatil.com/experience/industry/](https://www.indrapatil.com/experience/industry/)

I moved from academic research into software engineering to work closer to production systems: architectures, developer workflows, and domain applications that have to hold up under real constraints.

For the academic path that preceded this transition, see my [Research](/experience/research/) and [Education](/experience/education/) pages; related career and software-engineering talks are collected under [Presentations](/presentations/).

## Senior AI Engineer

Jan 2025 – Present

Munich, Germany

[Carl Zeiss AG](https://www.zeiss.com/)

-   **Owned the architecture and evolution** of the internal AI platform’s GenAI application layer — a shared **FastAPI/React/Terraform** template with software and agentic engineering standards and guardrails, adopted by **40 applications across business units**, delivering an estimated **€1.75M in productivity gains** through full-stack boilerplate reuse.
-   **Built LLM applications and agentic workflows:**
    -   **AI Launchpad**: company-wide system of record for AI initiatives (applications, MCPs, agents), used by the CFO and CEO to track AI maturity and strategic progress.
    -   **Expert debriefing**: searchable organisational memory from expert interviews (using **LangGraph** agents), mitigating institutional knowledge loss from expert attrition.
    -   **Medical-device coding agents**: custom quality harness for production-grade, regulation-aligned **C** code and legacy-modernisation agents — together delivering **35% efficiency gains and €1.3M in annual savings**.
-   **Drove enterprise-wide adoption of AI coding agents** (GitHub Copilot CLI, Codex, Claude Code), establishing standards and governance that enabled safe integration of **agentic coding workflows** across business units, and authored an internal _Agentic Engineering_ book (~1K visits/month).

## Senior Software Engineer

Oct 2023 – Dec 2024

Munich, Germany

[Preisenergie GmbH](https://www.preisenergie.de/)

-   **Architected and delivered** scalable browser-based **full-stack** applications (**FastAPI**/**Next.js**/**GCP**, initially **AWS**) serving the German energy market with high availability and performance requirements.
-   **Led infrastructure modernisation**, orchestrating microservices architecture in **Kubernetes** clusters through automated CI/CD pipelines (CircleCI, ArgoCD).
-   **Principal architect and technical lead** for **full-stack Shiny** applications ecosystem, **R package** development, and mission-critical **ETL pipeline** infrastructure, transforming previously unmaintainable legacy systems into well-structured, developer-friendly codebases.

## Software Engineer

Sep 2022 – Nov 2022

Berlin, Germany

[cynkra Analytics](https://cynkra.com/)

-   Worked on a **full-stack Shiny web application** in **R** for **financial data management** for the Swiss government.
-   **Improved and added GitHub Actions workflows** used by over **50 open-source projects**.

## Senior Research Software Engineer

Sep 2021 – Aug 2022

Berlin, Germany

[esqLABS GmbH](https://esqlabs.com/)

-   **Led technical implementation** of advanced PBPK modelling, simulation, and visualisation systems by architecting a scalable, **object-oriented** modular R package [ecosystem](https://github.com/orgs/Open-Systems-Pharmacology/repositories?language=r&type=all) to enhance long-term maintainability and extensibility.
-   **Drove feature development and quality assurance** for production **full-stack Shiny web applications** in the pharmacometrics domain, ensuring robust testing frameworks and production reliability.

## Open-Source Contributor

Jan 2018 – Present

Worldwide

[Independent and collaborative projects](https://github.com/IndrajeetPatil)

-   **Created and co-authored R and Python packages** for data analysis, statistical visualisation, reporting, and developer tooling, including **ggstatsplot**, **statsExpressions**, the **easystats** ecosystem, **lintr**, **styler**, **crosszip**, and **md2linkedin**. The [package portfolio](/experience/open-source/) collects the projects, authorship roles, and download counts.
-   **Built and maintained applications, documentation, and development tooling**, from a [full-stack chatbot template](https://github.com/IndrajeetPatil/chatbot-template) to this website, reproducible publications and presentations, and [reusable GitHub Actions workflows](https://github.com/IndrajeetPatil/workflows).

Continuity

My **open-source software development** began before my industry roles and continues alongside them. The [Open-Source](/experience/open-source/) page provides the **R and Python package portfolio** and cumulative download counts; the experience above describes the engineering work behind it.

## Technologies & tools

Technologies and engineering approaches from my work across research, open source, and industry.

Technology and engineering experience by area
| Area | Technologies, tools & approaches |
| --- | --- |
| Languages | Python, R, TypeScript, JavaScript, C, SQL, Bash, Zsh, PowerShell, HCL, KQL, CMake, reStructuredText |
| Full-stack frameworks | Next.js, Shiny (R), Astro, Django, NiceGUI |
| Backend development | Python, FastAPI, Node.js, Express, plumber, REST APIs, Django REST Framework, Pydantic, Server-Sent Events, OpenAPI |
| Frontend development | React, Preact, HTML, CSS, Tailwind CSS, Bootstrap, Material UI, Responsive design, Web accessibility, TanStack Query, React Router, React Hook Form, Zod, KaTeX, Recharts, styled-components, Markdown rendering (unified/remark/rehype) |
| Databases | PostgreSQL, SQLite, MongoDB, Microsoft SQL Server, Azure Cosmos DB (MongoDB API) |
| Cloud platforms | Microsoft Azure, Amazon Web Services (AWS), Google Cloud Platform (GCP), Azure Container Apps, Microsoft Entra ID, Azure Key Vault, Azure Database for PostgreSQL, Azure Storage, Log Analytics |
| Infrastructure | Terraform, Docker, Kubernetes, nginx, DNS management (Cloudflare), AzAPI, Private networking (VNets, private DNS), Managed identities, Multi-environment deployments, Docker Compose, Dev Containers |
| Version control & hosting | Git, GitHub, Bitbucket, Azure DevOps, Git worktrees, Conventional Commits, Semantic versioning, Changelog automation |
| Automation & deployment | GitHub Actions, CircleCI, Argo CD, GitHub Pages, Azure Pipelines, Workload identity federation, Reusable workflows, GitHub Apps, Multi-arch Docker builds, Release automation |
| AI & agents | Azure OpenAI, LangGraph, Model Context Protocol (MCP), LLM applications, Agentic workflows, LangChain, OpenAI Agents SDK, Vercel AI SDK, Anthropic Claude API, Agent observability (OpenInference), AI-assisted requirements and design generation |
| Coding agents | GitHub Copilot CLI, Codex, Claude Code, Antigravity CLI, AGENTS.md, Agent skills, Custom agents and prompts |
| Local model inference | Ollama, Local model hosting, Apple Silicon inference |
| Literate programming | LaTeX, Quarto, R Markdown, Jupyter, Markdown, marimo, Observable Framework |
| Documentation | MkDocs, Zensical, API documentation, Reproducible publications, Sphinx, Doxygen, Backstage TechDocs, Material for MkDocs, Link checking (lychee) |
| Document publishing | Pandoc, XeLaTeX, Typst, PDF and Word generation, Document conversion, python-docx, python-pptx, pdfmake |
| Presentations & diagrams | reveal.js, Mermaid, SVG |
| Statistical computing | Statistical modelling, Statistical visualisation, Bayesian and robust analyses, Meta-analysis, Repeated-measures analyses, Model diagnostics, Statistical reporting |
| Scientific & data workflows | Data preparation, ETL pipelines, Columnar data processing (Parquet), Schema validation, Financial time-series analysis, PBPK modelling, Simulation, Pharmacometrics, pandas, Polars, Plotly, Altair, Excel generation (XlsxWriter) |
| Software design | Object-oriented design, Functional programming, API design, Modular architecture, Microservices, R and Python package development, Command-line interface design, Plugin development, Code generation (Jinja2), Event-driven architecture, Asynchronous programming, Hardware abstraction, GUI tooling (NiceGUI, Tkinter) |
| Embedded systems | Zephyr RTOS, Embedded C, Kconfig, Devicetree, NXP i.MX RT1062 (Cortex-M7), CAN / CAN-FD, MQTT, J-Link, pyOCD, native\_sim |
| Requirements engineering | Polarion, Requirements traceability |
| Package publishing | CRAN, PyPI, GitHub Releases, Trusted publishing (OIDC) |
| Testing | Unit and integration testing, End-to-end testing, Load testing, Regression testing, Property-based testing, Mutation testing, Accessibility testing, Performance benchmarking, pytest, Vitest, Playwright, Testing Library, Locust, Snapshot testing, Ztest, Twister, Hardware-in-the-loop testing, Emulator-based testing |
| Development tools | GNU Make, just, uv, pnpm, Vite, Vite+, CMake, west, Task (Taskfile), Docker |
| Development environments | macOS, Windows, Homebrew, WinGet, Cross-platform setup automation, Linux, VS Code, Dev Containers |
| Search & discovery | Full-text search, Multilingual search, SEO, Structured data (Schema.org / JSON-LD), Agent-readable documentation (llms.txt) |
| Code quality | Static analysis, Type checking, Code formatting, Test coverage, Dependency audits, Ruff, mypy, ty, Pyrefly, ESLint, Biome, Prettier, clang-format, cppcheck, MISRA C, Axivion, Dead-code detection (vulture, knip, deptry), commitlint, gitlint, Commitizen, prek |
| Monitoring & security | Sentry, Lighthouse, Google Analytics, Google Search Console, Secret scanning, Workflow security, OpenTelemetry, Azure Monitor, Application Insights, Checkov, Trivy, gitleaks, CodeQL, Black Duck, Bandit, OAuth 2.0 / OIDC (MSAL), Licence compliance (SPDX) |
