Four skills to turn into a project
These are technology signals displayed in Opened’s catalogue, which can include alternatives, preferred tools, or tools already used by the team. We used the first page from each job-family filter on 17 November. Employer and seniority mix can skew those pages, so treat the numbers as a snapshot of these listings.
| If you want to work on | What showed up in our sample | A small project you can show |
|---|---|---|
| AI applications | Python appeared in 10 of 24 AI listings across three employers. Stripe accounted for eight of those mentions. | Build a Python service that answers questions using a small set of public documents. Show the sources it uses and what it says when the answer is missing. |
| Backend services | Java appeared in 16 of 24 backend listings across three employers. Kubernetes appeared in 11. | Build a Java API that records an event once, even if the caller retries. Include an integration test for the retry and explain how you chose the event key. |
| Frontend products | React appeared in 16 of 24 frontend listings across nine employers. TypeScript appeared in 12. | Build a keyboard-friendly filter UI with typed data and explicit loading, empty and error states. Record a short demo of each. |
| Data engineering and analytics | SQL appeared in 18 of 24 data listings across five employers. Python appeared in 16. | Load a public dataset, document the grain of each table, and write queries that catch duplicates and missing joins before they reach your dashboard. |
Sources: Python tutorial; Learn Java; React with TypeScript; PostgreSQL SQL tutorial; AI openings; Backend openings; Frontend openings; Data openings
Where LangChain fits
In a separate catalogue search, eight verified AI openings across six employers explicitly mentioned LangChain in their descriptions. Six grouped it with other tools or equivalent frameworks. One GitLab team described building reusable flows with LangChain and LangGraph. Another GitLab team listed hands-on experience with frameworks such as LangGraph or LangChain as a plus. In those descriptions, a candidate with equivalent experience has something concrete to talk about.
If the first row of the table fits your interests, start with one tool and a small set of examples. Include an answer supported by the source, a question the source cannot answer, and a case where the tool is unavailable. Put the expected behavior and the observed output in the README. That gives a reviewer something to evaluate when you describe the project.
Sources: Acceldata posting; GitLab Agent Foundations posting; GitLab Chat posting; LangChain Python quickstart; LangChain Academy
Trending on GitHub · Claude Quickstarts
GitHub’s weekly Python Trending page showed 1,241 stars this week for Anthropic’s Claude Quickstarts in our 17 November capture. Start with the customer-support agent example. Read how it connects the assistant to its knowledge base, then try the same pattern with three public help articles. Run one supported question and one deliberately unanswered question. If you run the full quickstart, it uses a Claude API key. The trend count measures GitHub interest. The employer postings above provide the hiring evidence.
Sources: GitHub Python Trending; Claude Quickstarts; Customer-support example
Put one example on your profile
If you’ve already used one of these tools, add the project link, what you personally built, and how you tested it. If you’re just starting, pick one project above and write down a test before adding the tool to your profile. A concrete example helps you explain the work when an employer asks.
Sources: Update your Opened profile