Backend systems for messy financial data

I’m Manikanta Nidadhavolu, a Python backend developer building OCR pipelines, LLM workflows and FastAPI services for international banks at PowerCred Technologies.

See my projectsEmail me

About me

I’m a Python backend engineer with 1.7 years of experience building document-processing systems for fintech. My work covers OCR ingestion pipelines, LLM-based classification and FastAPI services that handle millions of bank transaction records.

I like turning unreliable input, such as scanned statements, malformed dates and inconsistent formats, into data teams can trust. I also build the React dashboards that make that data easy to use.

I work remotely with PowerCred Technologies, a Singapore-based fintech company, and I’m open to backend and full stack engineering roles.

Role
Full Stack Developer
Company
PowerCred Technologies (remote)
Experience
1.7 years in fintech
Education
B.Tech in Information Technology
Looking for
Backend and full stack roles

Results at work

What I’ve delivered in 1.7 years of building fintech document-processing systems.

99.5%success rate on bank transaction pipelines
94%precision in transaction classification
95%PDF tamper-detection accuracy over 100K+ documents
70%less integration time with a standardization framework

What I do

The kinds of problems I can take on from start to finish.

OCR and document pipelines

Ingest scanned and unstructured files in 50+ formats, repair bad data and normalize it into one clean schema.

LLM and AI workflows

Add GPT-4, Gemini or Groq to your product for classification, RAG chatbots and async batch processing.

API development

Design fast, validated FastAPI services with Pydantic schemas, pagination and tests.

Fraud and validation

Detect tampered PDFs and suspicious transactions with rule-based checks and fuzzy matching.

Cloud and security

Migrate storage between GCP and AWS, and secure APIs with JWT, RBAC, sanitization and monitoring.

Dashboards

Build React and Next.js frontends that turn backend data into charts, search and filters.

Experience

Full Stack Developer, PowerCred Technologies

Singapore (Remote) · Mar 2025 – Present

Core backend and data

  • Built transaction pipelines for international banks that automate OCR ingestion at a 99.5% success rate.
  • Wrote a parsing engine for merged fields, malformed dates and nested records, cutting manual correction by 60%.
  • Designed adaptive architecture for 50+ unstructured formats and 1M+ records a month.
  • Used regex reconstruction to recover 20% more usable data from low-quality scans.

AI and fraud detection

  • Built an asyncio and GPT-4 system that processes million-row datasets 4x faster.
  • Improved classification to 94% precision with fuzzy matching and keyword extraction.
  • Launched a Fraud Indicator Module for mirror transactions and balance deviations, surfacing 85% more suspicious activity.
  • Validated PDF structure through metadata and object analysis, cutting false positives by 30%.

APIs and cloud

  • Shipped FastAPI endpoints with sub-200ms latency for 10k+ daily sessions.
  • Migrated storage from GCP to AWS S3 with 100% data integrity and zero downtime.
  • Secured 25+ APIs with JWT and RBAC across 6 roles.
  • Added Pydantic schemas and better pagination, shrinking payloads by 35%.

Product engineering

  • Led 10+ React dashboards on 5+ data sources for 100+ users.
  • Tuned data-fetching hooks, raising user engagement by 25%.
  • Standardized UI components, cutting new-feature frontend time by 30%.
  • Automated debugging workflows, saving 2 days per sprint.

Full Stack Developer Intern, Techplement

Maharashtra · Jun 2024 – Jul 2024

  • Co-developed Active Aura, a fitness tracker with 10+ modules and 200+ test users.
  • Built a React dashboard for 200+ daily activities and designed a MongoDB schema for 500+ user profiles.
  • Improved performance by 30% and made 100+ Git commits with a 5-person team.

Projects

Two full-stack builds that combine backend logic with LLMs.

InsureIQ

Insurance document intelligence and claim risk assessment, from upload to risk score at 95% accuracy.

  • OCR and field extraction with PyMuPDF and an LLM, sorting documents into 8 categories.
  • Rule-based risk engine that scores 0–100 across 4 levels and explains each factor.
  • 3-tier validation of fields, business rules and cross-document consistency.
  • 15+ REST endpoints, a dashboard with charts and search, and a Pytest suite.
FastAPIReactTypeScriptSQLitePyMuPDFPytest

View on GitHub

AI Recruitment Copilot

A job-application assistant with 7+ features, from ATS scoring to recruiter outreach.

  • ATS engine with 5 factors, including semantic similarity and experience parsing.
  • RAG career chatbot on Qdrant and Groq with 4 query routes.
  • Generators for cover letters, interview prep and 30-day learning roadmaps.
  • Bulk Gmail outreach with configurable delay and an 8+ endpoint API.
FastAPINext.jsQdrantGroqsentence-transformers

View on GitHub

Skills

The tools I use to build and ship.

Languages

PythonJavaScriptTypeScriptSQL

Backend

FastAPIREST APIsMicroservicesPydanticAsyncioJWT/RBACPytest

AI and data

GPT-4GeminiGroqPrompt engineeringRAGQdrantOCRETLRegexFuzzy matching

Frontend

React.jsNext.jsAngularHTMLCSS

Cloud and DevOps

AWS S3AWS TextractGCPAzureDockerCI/CDGit

Databases

PostgreSQLMongoDBBigQuerySQLite

Education and certifications

 

B.Tech in Information Technology

SRKR Engineering College, Bhimavaram · Aug 2019 – Apr 2023 · CGPA 8.67/10

Data StructuresOOPDatabasesOperating SystemsNetworksMachine Learning

Certifications

  • The Complete JavaScript Course 2025, Udemy
  • Cloud Computing, Coursera
  • Python Crash Course, Coursera

Let’s work together

I’m looking for backend and full stack engineering roles.

manikantanidadhavolu@gmail.com +91 73968 29004 LinkedIn