BACK TO EXPERIENCE
FORMER INTERN / 2026

Software Engineering
Internship.

Accenture Solutions Pvt. Ltd.

Built the backend for an AI-assisted Pull Request review system that turns verified webhooks and exact-commit code context into deterministic baseline reviews before LLM handoff.

18 May — 10 July 20268 WEEKS
Explore the work
01 BACKEND SYSTEMS
01 / WHAT I BUILT

From event
to review.

A reliable processing pipeline that assembles precise code context before review generation.

01

Verified webhook ingestion

Built FastAPI webhook endpoints with HMAC-SHA256 payload verification.

02

Precise code context

Computed Git diffs against exact commit SHAs and extracted structured code context with Python AST.

03

Reliable async processing

Implemented Celery and Redis processing with delivery idempotency checks.

04

Modular review engine

Combined deterministic heuristics with contextual knowledge retrieval to prepare baseline reviews before LLM handoff.

02 / EIGHT WEEK TIMELINE

Build, verify,
hand over.

Eight weeks from architecture discovery to a documented project handover.

WEEKS 01–03

Onboarding & architecture

Knowledge transfer, system discovery, and architecture design.

WEEK 04

Webhook ingestion

Implemented verified webhook endpoints and the initial event flow.

WEEK 05

Diffs & code context

Built exact-SHA Git diff computation and Python AST context extraction.

WEEK 06

Async reliability

Added Celery and Redis processing with delivery idempotency checks.

WEEK 07

Review intelligence

Developed heuristic reviews and contextual knowledge retrieval.

WEEK 08

Presentation & handover

Presented the completed system and documented the handover.

03 / TECHNOLOGY
PythonFastAPICeleryRedisPython ASTGitHMAC-SHA256
Back to portfolioLet's talk