Ask HN: What tools or processes accelerate engineer onboarding?

1 point

7 months ago

I recall onboarding a US-based software engineer in 2018-19 was estimated at ~$80–120K (salary, lost productivity, and ramp time). A recent research publication from DX (AI-Assisted Engineering: Q4 Impact Report) [1] offers updated data that reframes how we think about onboarding.

Here are three relevant facts from the report:

1. Onboarding speed improvement:

  Engineers using AI coding assistants daily 
  reached their 10th merged pull request 
  in about 49 days, compared to about 
  91 days for engineers not using 
  new AI-powered tools.

2. Time savings per engineer:

  Developers using AI tools report saving 
  about 3.6 hours per week.

3. High adoption of AI tools

  Over 90% of developers across the 
  sample use AI tools at least monthly.

Curious how y'all feel about this info from where you're sitting.

Few immediate caveats:

  - I definitely feel that DX has a biased sample 
    because eng orgs that use them are surely forward
    thinking and the leadership value developer
    productivity.
   
  - The 10th merged PR is a DX metric, 
    so it's possible that's not the best or 
    there's a better metric design.

Questions for the community:

1. Call the BS, what's not true in your world?

2. What tools & processes do you swear by, what have you found that doesn't move the needle?

3. Is Onboarding even a problem to focus on and what's a definition of good enough for that process?

---

References

1: DX Report: https://getdx.com/uploads/ai-assisted-engineering-q4-impact-report.pdf