ANIKET.
Architecting production-grade AI systems, ultra-fast async backends, and multimodal signal processing pipelines.
Engineering Experience
FinProhub
AI Finance Tracker PlatformPRODUCTION ACTIVE- ➢Tenforty SystemIntegrated Stripe Payment Gateway for PCI-compliant checkout sessions, recurring subscription lifecycle management, and automated invoice processing.
- ➢Tenforty SystemEngineered zero-downtime Stripe Webhook listeners (checkout.session.completed, invoice.payment_succeeded) with idempotent database event processing and retry workflows.
- ➢FinProhub CoreIntegrated Azure OCR + PaddleOCR fallback with Pydantic validation, reducing document processing latency from 15s down to 3s.
- ➢FinProhub CoreBuilt RAG-powered chatbot for financial queries using vector embeddings and LLM generation.
- ➢FinProhub CoreBuilt and shipped JWT-based authentication securing user sessions, REST APIs, and automated GoDaddy SMTP notification pipeline.
B.E. in Artificial Intelligence and Machine Learning
VIDBRAIN
AI pipeline that converts long-form videos into concise, topic-focused summaries with transcription, HDBSCAN semantic clustering, and vector search.
3D MOTION ANALYSIS
Markerless motion capture system analyzing athlete biomechanics from standard video, providing a cost-effective alternative to sensor-based hardware.
AURA
Adaptive AI assistant engineered for multi-turn reasoning, dynamic context selection, and automated task execution across heterogeneous datasets.
FIELD ENCRYPTOR
Lightweight, high-performance security utility for granular field-level encryption and secure payload transmission across distributed backend services.
AUDIO SUBTITLE EXTRACTOR
Automated signal processing utility designed to extract, segment, and align audio streams and subtitle metadata from complex multi-track video containers.
I don't start with the model.I start with the system.
PROBLEM
Define what needs to be solved, not what model to use
DATA
Understand the signal before choosing representation
REPRESENTATION
Embeddings, features, structured extraction
MODEL
Choose the right tool, not the trending one
EVALUATION
Measure what matters to the system
INFRASTRUCTURE
APIs, databases, authentication, monitoring
DEPLOYMENT
Ship it. Make it reliable.
Technical Ecosystem
Hover any technology to reveal interconnected AI pipeline & infrastructure dependencies.
AI / ML
Computer Vision / OCR
Backend / Infrastructure
Databases
"I bridge the gap between complex AI research and reliable production systems."
I am an AI Systems & Backend Infrastructure Engineer specialized in building resilient APIs, real-time multimodal processing pipelines, and vector-backed intelligence applications.
1. Production AI & RAG Infrastructure
Architecting vector search databases (FAISS), hybrid BM25 + dense retrieval, and low-latency LLM agent pipelines with zero-hallucination guardrails.
2. High-Throughput REST & Payment Backends
Building zero-downtime Stripe payment integrations, webhook listener idempotency, JWT session security, and FastAPI microservices.
3. Multimodal Signal & OCR Optimization
Deploying Azure OCR + PaddleOCR computer vision fallback pipelines with Pydantic validation, optimizing document extraction latency by up to 80%.