Economics Engineering Student → Data Engineering
Build.Measure.Verify.
Discipline
Economics × softwareMethod
Small slices, testedOutput
Systems you can open
I turn economic data into tested, reproducible systems with Python and SQL—then expose the evidence, limits, and engineering decisions behind every claim.
Open to data engineering internships and junior opportunitiesBursa / TürkiyeUTC+03
Python · SQL · Econometrics · Data Engineering · Simulation · VerificationPython · SQL · Econometrics · Data Engineering · Simulation · Verification
Tested > claimed · Deployed > demoed · Evidence > adjectivesTested > claimed · Deployed > demoed · Evidence > adjectives
01 / Selected systems
Work with
receipts.
Four focused systems. Each card states my ownership, the key engineering choice, the measured result, and the boundary between tested and deployed.
01 — FLAGSHIPEnflasyonumdan ne haber?
A personal inflation engine for households whose spending basket differs from the official average.
My ownership: Designed and implemented the calculation pipeline, API, persistence, tests, CI, and deployment.
Key decision: Used a Laspeyres index across 13 ECOICOP sub-indices so results remain explainable and comparable with official CPI.
13ECOICOP sub-indices
Laspeyresindex method
3.11+Python requirement
PythonFastAPIPostgreSQLCI
Live vertical slice · free host may need 30–60 seconds to wake
02 — LEARNING SYSTEMMakroQuest
An evidence-based economic detective game that turns macroeconomic reasoning into an inspectable scenario.
My ownership: Built the first end-to-end case, sourced retrieval, RAG hint agent, evaluation, FastAPI interface, and deployment.
Key decision: Kept citations and evaluation beside generation so the learning experience does not hide unsupported answers.
1implemented case
RAGsourced retrieval
Weeklylive smoke workflow
FastAPIRAG evaluationWorld BankDocker
M1 deployed · free host may need 30–60 seconds to wake
03 — DATA PLATFORMecon-lakehouse
A medallion lakehouse for Turkish macro data, from validated Bronze Parquet to analytical marts, with a public run-evidence page.
My ownership: Built ingestion, validation, dbt/DuckDB models, API, dashboard, orchestration, run-audit ledger, and CI.
Key decision: Separated Bronze, Silver, and Gold contracts to make lineage, data quality, and failures observable.
05:23 UTCdaily scheduled run
fixturesynthetic data mode
30hstale-after threshold
dbtDuckDBParquetDagster
Tested, not production-ready · runs on synthetic fixtures, freshness bounded
04 — SIMULATION R&DHomefront Universe
A zero-runtime-dependency WebGL2 space-RTS engine with hand-written GLSL and a headless deterministic simulation path.
My ownership: Implemented the renderer, shaders, simulation loop, headless verification, and immutable proof artifacts.
Key decision: Stamped committed frames with engine-state checksums so visual results are reproducible rather than decorative.
1,337documented seed
3,000verified ticks
985466095reference checksum
WebGL2GLSLSimulationDeterminism
Deterministic evidence · commit-pinned reproduction path
02 / Live surfaces
Open them.
They run.
Static-hosted builds that need no cold start. Each one opens in your browser and shows its own method, source, and limits.
03 / Verification
Claims are
not proof.
The evidence ledger ties public statements to immutable revisions and records limitations instead of hiding them.
Open evidence ledger →- Immutable 40-character commit references
- Explicit tested versus deployed boundaries
- Known hosting and operational limitations
- No unsupported performance claims
04 / About
Economics,
made executable.
I study Economics Engineering and build toward data engineering—connecting economic reasoning, reproducible pipelines, software quality, and simulation.
I optimize for small vertical slices, automated tests, explicit boundaries, and evidence that a system actually runs—not technology lists or unverified claims.