Andrei Bogomolov
Python Scripts & Data Automation
New to Contra
Andrei is ready for their next project!
File handoff verification with SHA-256 manifests
AI-assisted, self-initiated Python demonstration using invented text files. A read-only scanner records relative paths, byte counts and SHA-256 hashes; a comparison function reports added, removed, changed and unchanged files.
Verified sample: one added file, one removed file, one changed file and one unchanged file across two synthetic snapshots. A known SHA-256 test vector and content-preservation check passed.
Outputs: two JSON manifests and a categorized difference report. Symlinks are outside scope and rejected. This is an integrity check, not malware scanning or a production backup system. Part of a three-tool suite with six passing tests. Source and synthetic samples available on request; no prior client work is claimed.
0
7
Unicode label cleanup with a collision report
AI-assisted, self-initiated Python demo using seven invented labels. Applies Unicode NFC and folds whitespace while preserving case, record order and string identifiers.
Verified result: four labels changed, one collision group retained for review, and one empty result flagged. Nonbreaking spaces and tabs become ordinary single spaces. 0010 and 10 stay distinct. Running the cleaned labels again produces no further changes.
Outputs: normalized labels, before/after log, collision list and empty-record list. Part of a three-tool suite with six passing automated tests. Source and synthetic inputs available on request. No client work or production deployment is claimed.
0
11
JSON export validation before import
AI-assisted, self-initiated Python demonstration on synthetic data. Validates a small customer-export schema: nonempty string IDs, booleans and real YYYY-MM-DD dates. Reports duplicate IDs and malformed values by record and field without modifying the source.
Verified example: five records, six issues across three records. IDs 0010 and 10 remain different; 2024-02-29 passes, 2026-02-30 is flagged.
Included in a three-tool demo suite with six automated tests passing. Source, test data and result JSON available on request. This is a fixed demo schema, not a production deployment or a past client project.
0
15
CSV cleanup with a traceable change report
AI-assisted, self-initiated demonstration using synthetic records—not a client project. A working Python utility applies agreed whitespace/case rules, removes exact full-row duplicates, and flags uncertain records without guessing.
Verified sample: 12 input records → 11 retained + 1 duplicate; 7 cells normalized; 6 review flags. IDs 0010 and 10 stay distinct. The original file is preserved byte for byte.
Outputs include cleaned data, cell and record logs, a review list, and reconciled counts. Five integrity checks and seven edge-case tests passed. Source and synthetic sample are available on request.
0
19