Software engineer.
AI agents are changing how software gets built, now and going forward β but a human still has to validate the work, judge the architecture, decide what to build, why, and when to stop, and just as importantly, what not to build.
Principles
- Changeability over cleverness
- Understand before changing
- Simplicity through removal, not addition
- Evidence over opinion
Currently exploring (free-time, need- and interest-driven, not a curriculum)
- Functional programming
- Domain-driven design
- Distributed systems
- AI agentic workflows
- Electronics & Arduino
- Mechanical engineering
- Motorcycle racing training
Sinh ra rα»i cΕ©ng sαΊ½ tΓ n,
CΓ΄ng danh nhΖ° bα»t trΓͺn ngΓ n sΓ³ng xa.
ThΓ’n nΓ y gΓ‘ tαΊ‘m phΓΉ hoa,
TrΔm nΔm rα»i cΕ©ng hoΓ‘ ra hΖ° khΓ΄ng.
NguyΓͺn tαΊ―c nhΖ° mα»t chuyαΊΏn ΔΓ²,
Qua sΓ΄ng rα»i bα», ai Δo bαΊΏn bα».
Γm khΖ° mα»t mα»i tΓ΄n thα»,
Tay ΔΓ’u mΓ mα» lα»i chα» ngΖ°α»i sau.
Trang nΓ y cΕ©ng chα» phΓΉ vΓ’n,
ΔΓ΄i dΓ²ng chα»― mα»n cΓ³ ngαΊ§n αΊ₯y thΓ΄i.
Mai kia giΓ³ cuα»n xa xΓ΄i,
Chα»― tan theo giΓ³, ngΖ°α»i trΓ΄i phαΊn mΓ¬nh.
NgΖ°α»i Δang Δα»c mαΊ₯y dΓ²ng ΔΓ’y,
Mai nΓ y cΕ©ng chαΊ³ng nhα» ngΓ y ΔΓ£ qua.
Lα»i tΓ΄i nhαΊΉ tα»±a sΖ°Ζ‘ng xa,
Nghe rα»i thΓ¬ bα», ΔΖ°α»ng xa tα»± tΓ¬m.
NΓ³i buΓ΄ng mΓ dαΊ‘ chΖ°a buΓ΄ng,
Khoe cΓ’u giΓ‘c ngα» cΕ©ng tuα»ng vαΊ½ vα»i.
Bα»n cΓ’u nΓ y cΕ©ng lΓ lα»i,
Δα»c xong rα»i bα», mα»t Δα»i nhαΊΉ tΓͺnh.
π± My GitHub Data
π¦ 366.6 kB Used in GitHub's Storage
π 42 Contributions in the Year 2026
πΌ Opted to Hire
π 341 Public Repositories
π 21 Private Repositories
I'm an Early π€
π Morning 1197 commits βββββββββββββββββββββββββ 17.99 %
π Daytime 2211 commits βββββββββββββββββββββββββ 33.24 %
π Evening 2634 commits βββββββββββββββββββββββββ 39.60 %
π Night 610 commits βββββββββββββββββββββββββ 09.17 %
π I'm Most Productive on Tuesday
Monday 701 commits βββββββββββββββββββββββββ 10.54 %
Tuesday 1669 commits βββββββββββββββββββββββββ 25.09 %
Wednesday 917 commits βββββββββββββββββββββββββ 13.79 %
Thursday 530 commits βββββββββββββββββββββββββ 07.97 %
Friday 867 commits βββββββββββββββββββββββββ 13.03 %
Saturday 1148 commits βββββββββββββββββββββββββ 17.26 %
Sunday 820 commits βββββββββββββββββββββββββ 12.33 %
π This Week I Spent My Time On
π¬ Programming Languages:
Markdown 17 hrs 4 mins βββββββββββββββββββββββββ 70.18 %
Elm 2 hrs 34 mins βββββββββββββββββββββββββ 10.58 %
Ruby 1 hr 54 mins βββββββββββββββββββββββββ 07.87 %
HTML 1 hr 16 mins βββββββββββββββββββββββββ 05.25 %
YAML 53 mins βββββββββββββββββββββββββ 03.65 %
π₯ Editors:
Claude Code 22 hrs 44 mins βββββββββββββββββββββββββ 93.45 %
Neovim 1 hr 35 mins βββββββββββββββββββββββββ 06.55 %
π€ AI Coding This Week
β± AI Coding Time: 23 hrs 40 mins (97.3%)
βοΈ 9,544 lines written by AI, 418 lines written by hand (95.8% AI-written)
π€ 316,795,991 Input Tokens, 966,495 Output Tokens
π΅ $642.78 Estimated AI Cost This Week
π§ 27 AI Sessions, 790 AI Prompts
Sonnet 6,445 lines βββββββββββββββββββββββββ 61.18 %
Claude 4,090 lines βββββββββββββββββββββββββ 38.82 %
π AI Coding Insights:
π€ AI-Driven β 95.8% of written lines came from AI
π Verbose Prompter β average 2,109 characters per prompt
π Iterative Prompter β average 29 prompts per session
π High AI Trust β 4.63% of changed lines were hand-edited
I Mostly Code in Ruby
Go 22 repos βββββββββββββββββββββββββ 19.30 %
Haskell 4 repos βββββββββββββββββββββββββ 03.51 %
Clojure 2 repos βββββββββββββββββββββββββ 01.75 %
Elm 1 repo βββββββββββββββββββββββββ 00.88 %
TypeScript 1 repo βββββββββββββββββββββββββ 00.88 %
Timeline
Last Updated on 20/08/2026 16:28:06 UTC