// adequately-engineered

Built for
the annoying
problems.

Software for problems that were real, genuinely irritating, and fixable with a boring stack. No LLMs doing arithmetic. No Kubernetes for a side project. No blockchain anywhere. The thing works. We're done here.

projects — 3 shipped · 1 in progress
001 · tax

IndITR

"The Tax Agent That Doesn't Let AI Do The Math"
Every July: Form 16 in the inbox, Zerodha P&L downloaded, IT portal timing out, no idea if that midcap sale is STCG or LTCG. You build distributed systems. You cannot file your own taxes.

IndITR is a conversational agent for AY 2026-27. It reads your documents — Form 16, Zerodha/Upstox P&L, bank statements — and computes your ITR-1 or ITR-2 in pure Python. LangGraph runs the conversation. Python does every rupee of arithmetic. Old regime vs new. 80C deductions. HRA. Capital gains with the new rules. 301 tests make sure it stays correct. Data never leaves your machine.
301
tests passing
0
LLM arithmetic calls
2
ITR forms
LangGraph Ollama Python Pure-Python Tax Engine AY 2026-27 Qwen 9B
acervenky/adequately-engineered-inditr
002 · jobs

JobMemo

"Know If You're a Fit. Know If They're Worth It. Then Apply."
Tab-switching hell: paste JD into Claude, Google the company, check Glassdoor, repeat 20 times a day with no memory of what you decided about the last one. Genuinely awful.

JobMemo is a Chrome extension that lives directly on the job portal. It scores the role against your resume — 0 to 100, temperature 0, same rubric every time, so the same job always scores the same and it’s not a mood. It researches the company with a live agentic search loop. It writes your cover letter in context. No backend. No account. Nothing leaves your browser. Run it on Ollama and nothing leaves your machine at all.
9
portals supported
0
backends
6-turn
research agent
Chrome MV3 Ollama JavaScript OpenRouter ReAct Agent No Backend
acervenky/adequately-engineered-jobmemo
003 · water

MumbaiDams

"All 7 Reservoirs. One Page. No Guesswork About Water Cuts."
Every pre-monsoon: five articles saying Mumbai will run dry, five saying it won’t. BMC publishes weekly storage numbers for each lake separately. Nobody puts them together. Nobody tells you whether the ENSO index this year actually implies a bad monsoon. Nobody models what that means for next October.

MumbaiDams pulls storage data from BMC, FPJ, and numerical.co.in, cross-validates against CWC official bulletins, layers in Open-Meteo rainfall corrected for the 38% peak-monsoon underestimate against IMD ground stations, tracks NOAA’s ENSO index, and runs 12-month projections. Tansa, Vihar, Tulsi, Modak Sagar, Bhatsa, Upper Vaitarna, Middle Vaitarna — all on one page. Static HTML. No server. No account.
7
reservoirs tracked
12mo
projections
0
backends
Vanilla JS Open-Meteo NOAA CPC IMD / CWC GitHub Pages No Backend
acervenky/mumbai-dam-tracker acervenky.github.io/mumbai-dam-tracker
004 · ???
Something Real
"found another problem. currently annoyed by it."
Identified something irritating. Confirmed it is not just me. Stack selection in progress — boring options only.

details redacted until it ships and actually works

ETA: when it’s done. No promises.
???
tbd
0
lines written yet
1
problem confirmed real
boring stack (probably) no kubernetes ships when it ships