$ whoami
✓ aditya_bhattacharjee — backend & data systems
$ status --check
✓ 6 services under active development · 3 inter-service integrations wired

Building backend systems
that talk to each other.

PYTHON DEVELOPER — BACKEND SYSTEMS, REST APIs, DATA PIPELINES

Third-year CSE student building a suite of six interconnected services with FastAPI, PostgreSQL, Redis, and Apache Airflow — designed as one system, not six separate projects, with CI/CD via GitHub Actions on every service.

View the system → Get in touch
6/6
Services in system
8.0
CGPA · Year II
3
Inter-service integrations
3
Certifications
System Design

One system, six services.

These aren't six standalone projects — they're designed to depend on each other, the way production systems do. Click a node to jump to its entry in the service registry below.

service-account JWT post-payment jobs click events · pub/sub independent ETL AuthCore JWT · RBAC · Redis blacklist FastAPI · PostgreSQL PayBridge Stripe webhooks · idempotency FastAPI · Stripe API LinkForge URL shortener · analytics FastAPI · Redis cache WorkerMesh Async task queue · Celery Priority queues · dead-letter FastAPI · Redis broker StreamPulse WebSockets · Redis pub/sub Real-time broadcast DataFlow ETL orchestration Apache Airflow · DAGs
Wired integration (resume-verified)
Service online / in active dev
Service Registry

Six services, one registry.

All under active development. Each entry below is what's actually implemented — expand for the technical detail.

01 LinkForgeURL Shortener & Click Analytics API Active Dev FastAPI · PostgreSQL · Redis · Docker ▾
FastAPI · PostgreSQL · Redis · Docker · GitHub Actions
  • Two-tier storage: Redis as the hot-path cache for O(1) redirect lookups, PostgreSQL as source of truth for link metadata and click events — separating read-latency from write-durability concerns.
  • TTL-based and event-driven cache invalidation, token-bucket rate limiting per API key (not per IP), and cursor-based pagination replacing OFFSET to avoid deep-page degradation.
  • Emits structured click events to a Redis Pub/Sub channel consumed by StreamPulse, driving real-time analytics without direct service coupling.
  • GitHub Actions runs pytest + flake8 on every push, builds and tags a Docker image on success, deploys to staging on merge to main.
02 AuthCoreJWT Identity & Access Management API Active Dev FastAPI · PostgreSQL · Redis · bcrypt ▾
FastAPI · PostgreSQL · Redis · bcrypt · Docker
  • Full JWT lifecycle: short-TTL access tokens with long-TTL refresh tokens stored bcrypt-hashed in PostgreSQL, rotated automatically on every refresh to narrow replay windows.
  • Stateless revocation with no per-request DB lookup — invalidated tokens land in a Redis blacklist keyed by JTI, TTL matched to remaining token validity.
  • RBAC across three actor types (admin, user, service account), permission scopes encoded in token payloads, FastAPI dependency injection for route-level authorization.
  • Exposes a service-account token flow consumed by WorkerMesh for inter-service authentication.
03 PayBridgeStripe Webhook Processing & Payment API Active Dev FastAPI · Stripe API · PostgreSQL · Redis ▾
FastAPI · Stripe API · PostgreSQL · Redis · Docker
  • Every incoming Stripe webhook is validated with HMAC-SHA256 signature verification before any DB interaction; Redis-backed idempotency keys stop retried requests from reprocessing.
  • Payment state modeled as a finite state machine (Pending → Processing → Succeeded / Failed / Refunded) with PostgreSQL row-level locking to prevent race conditions on concurrent transitions.
  • Exponential backoff retries for failed webhook deliveries, dead-letter log for exhausted-retry events, full audit trail for finance.
  • Dispatches post-payment jobs (confirmation emails, inventory updates) to WorkerMesh via task queue — keeps the payment API non-blocking.
04 WorkerMeshAsync Task Queue & Job Scheduling Service Active Dev FastAPI · Celery · Redis · PostgreSQL ▾
FastAPI · Celery · Redis · PostgreSQL · Docker
  • API accepts a job payload, persists a job record to PostgreSQL, dispatches to a Celery worker via Redis broker — returns a task ID immediately, caller is never blocked.
  • Priority queues with independent worker pools: user-triggered exports route high-priority, digest emails and cleanup route low-priority.
  • Authenticates inbound service-account requests through AuthCore, receives post-payment jobs dispatched by PayBridge — completing the coordination chain.
  • Task lifecycle (Queued → Running → Succeeded / Failed) tracked in PostgreSQL with structured logs; retry-exhausted tasks move to a dead-letter queue with full error trace.
05 DataFlowETL Pipeline Orchestration Engine Active Dev Apache Airflow · Python · PostgreSQL ▾
Apache Airflow · Python · PostgreSQL · Docker · GitHub Actions
  • Orchestrates a multi-source ETL pipeline via Airflow DAGs — ingests from a REST API, a scheduled CSV drop, and a source PostgreSQL DB, transforms in a Python layer, validates before loading to target schema.
  • Data quality gates at each stage: null-rate checks, schema validation, anomaly detection — tasks fail explicitly and alert rather than passing bad data downstream silently.
  • Retry logic, SLA monitoring, task-level structured logging per DAG — every run fully observable from the Airflow UI.
  • GitHub Actions lints DAG files, runs unit tests on transformation logic, deploys updated DAGs to production on merge to main.
06 StreamPulseReal-Time Event Broadcast Service Active Dev FastAPI · WebSockets · Redis Pub/Sub ▾
FastAPI · WebSockets · Redis Pub/Sub · PostgreSQL · Docker
  • Pub/sub architecture: FastAPI WebSocket handlers subscribe to Redis channels on connection, a publisher writes events, clients receive broadcasts within milliseconds — replaces polling with persistent push.
  • Explicit WebSocket lifecycle handling: heartbeat pings detect stale connections, clean disconnects unsubscribe, in-memory registry prevents phantom deliveries.
  • Publishes all events to PostgreSQL as an append-only log — new connections replay the last N events on join, no separate polling endpoint needed for history.
  • Subscribes to LinkForge's click-event channel, broadcasting live analytics updates — closes the loop from link creation to live reporting.
Dependencies

The stack.

What every service in the registry above is actually built with.

stack.yaml
languages:
Python, SQL, JavaScript, Java, C/C++, HTML
algorithms_and_ds:
Arrays, Hash Maps, Trees, Graphs, Dynamic Programming,
Binary Search, Sliding Window, Recursion
backend_and_apis:
FastAPI, REST API Design, JWT, OAuth2, WebSockets, Async Python
data_pipelines:
Apache Airflow, ETL/ELT, Batch Processing, Data Validation
databases:
PostgreSQL, MySQL, Redis # caching, pub/sub, token blacklisting
infra_and_cicd:
Docker, GitHub Actions, Git, Linux/Bash, Postman
architecture_patterns:
Celery, RBAC, Event-Driven Design,
Cursor Pagination, Idempotency
About

Currently building, currently learning.

I'm Aditya Bhattacharjee, a second-year B.Tech CSE student at SRM Institute of Science and Technology, specializing in backend systems, REST APIs, and data pipeline engineering. I'm building the six-service system above as one interconnected whole rather than isolated coursework projects — with performance, reliability, and observability treated as first-class requirements, not afterthoughts. I work comfortably independently in remote environments and practice CI/CD-first development with GitHub Actions on every service.

AI/ML Intern — DSLM Build & Deployment
Tech Mahindra Makers Lab — Mentor: Mr. Rama Krishna Prasad Koya
June 2026 — Present · Remote
  • Working under mentorship on the DSLM (Domain-Specific Model) track, building foundational understanding of the model development lifecycle.
  • Built exploratory mini-projects applying NLP, LangChain, and Retrieval-Augmented Generation (RAG) to gain hands-on exposure to LLM-based application patterns.
  • Studied core deep learning architectures, including Transformers and Convolutional Neural Networks, through guided implementation exercises.
  • Began exploring early concepts in agentic AI systems as part of ongoing skill development.
Frontend Development Intern
C.E. Info Systems Ltd. — Mappls / MapMyIndia
May 2026 — July 2026 · New Delhi, India
  • Assisted the engineering team in building and integrating frontend components for the revamped Mappls website and companion apps, working closely with a mentor to translate feature requirements into production code.
  • Debugged and resolved issues on the existing production website — including broken links, pages failing to render correctly, and links misrouting to incorrect destinations — improving site reliability ahead of the revamp rollout.
  • Contributed to integration between new frontend features and existing platform services as part of the broader website and app revamp.
  • Gained mentor-guided, hands-on experience in enterprise-scale frontend delivery for a nationally-used geospatial and mapping platform.
Python Research Intern
Dr. R. Geetha, Faculty of Artificial Intelligence — SRM IST
February 2026 — May 2026 · Chennai, India
  • Built Python utility scripts for data ingestion, preprocessing, and experiment logging across active research projects — emphasizing modularity and reusability.
  • Contributed to a Git-based collaborative codebase: inline documentation, modular design patterns, code review cycles with faculty and peers.
  • Translated researcher requirements into working Python implementations across multi-module projects, practicing the full spec-to-delivery cycle.
  • Maintained reproducible environments through standardized configuration and structured logging, keeping experimental results traceable.
Academic Standing
8.0
CGPA · Year II · B.Tech CSE
SRM Institute of Science and Technology
SRM Institute of Science and Technology B.Tech, Computer Science & Engineering · Expected March 2028 · Kattankulathur, Chennai
GD Goenka Public School, Gurgaon CBSE Class XII · April 2024 · 70%
Certifications
  • SQL for Data AnalysisIIT Roorkee via Intellipaat · Dec 2025
  • Artificial IntelligenceIIT Kharagpur · Kshitij
  • Object Oriented ProgrammingNPTEL · April 2025
Achievements & Leadership
  • 1st Place — Essentials of DevOps Workshop
  • 3rd Place — Alexaverse Webslinger Event
  • Professional Acknowledgement — IBM AI Intelligent Workshop
  • Class Representative (Year I) · IEEE Student Chapter & CSI · Red Bull Basement Ideathon
Open to internship opportunities

Let's talk backend.

Reach out about internships, collaboration, or the system above — happy to walk through any part of it.

Emailb.aditya2801@gmail.com
Phone+91 74001 52325
BasedKattankulathur, Chennai · India
GitHub
@fictus999
→
LinkedIn
Aditya Bhattacharjee
→
LeetCode
@Fictus999
→