Avinash Vikram Singh
Data Scientist · AI Engineer · Full-Stack Developer

Crafting intelligence from raw data

Currently MSc Data Science Christ University, Bengaluru
Experience 2+ Years ML, NLP, Cloud AI
Location Bengaluru India
Python/ TypeScript/ TensorFlow/ LangChain/ React/ AWS/ PyTorch/ Next.js/ GCP/ Transformers/ Docker/ Python/ TypeScript/ TensorFlow/ LangChain/ React/ AWS/ PyTorch/ Next.js/ GCP/ Transformers/ Docker/

About me

I build end-to-end intelligent systems — from conversational Voice AI and multilingual NLP pipelines to full-stack SaaS products and predictive ML engines. With a BSc from Delhi University and an MSc at Christ University, I work at the intersection of research and production.

I've deployed omnichannel AI agents for healthcare triage, engineered multi-tenant backends with real-time analytics, built RAG pipelines over large-scale vector stores, and shipped full-stack products used in production — all with a focus on latency, scalability, and measurable impact. I'm particularly interested in building AI tools that serve underserved populations, including low-resource language communities and high-stakes domains like healthcare and agriculture.

0 years exp.
0 products shipped
0 + languages supported

Work experience

AI Engineer Intern

PorsAI
Dec 2025 — Present Remote
  • Developed and deployed multilingual Voice AI and omnichannel chatbots (WhatsApp, Instagram, Web) for automated patient triage, appointment scheduling, and lead management
  • Engineered core backend for a scalable multi-tenant architecture, integrating multiple databases and building automations to streamline healthcare operations
  • Built a specialized clinical AI diagnostic application for hair and skin analysis, and implemented automated lead-scoring using predictive analytics to optimize conversion rates
Voice AILangChainMulti-tenantPredictive Analytics

Python (AI) Developer Intern

Indian Capital and Investment
Jun 2025 — Nov 2025 Remote
  • Orchestrated scalable ML environments with GCP and Vertex AI for model deployment
  • Built high-performance Python scripts for automated unstructured data extraction
  • Integrated AWS cloud-native tools, improving infrastructure reliability by 25%
GCPVertex AIAWSPython

Data Scientist

ARD Information Systems
Nov 2023 — May 2025 Noida, India
  • Developed Random Forest ranking algorithm, cutting screening time by 40%
  • Led EDA on a 50GB+ database, boosting placement success by 15%
  • Integrated ML ranking into ATS via REST APIs with <30ms latency
Random ForestREST APIsSQLEDA

Data Scientist Intern

ARD Information Systems
Jul 2023 — Sep 2023 Noida, India
  • Optimized complex SQL queries, reducing recruiter profile retrieval latency
  • Refactored legacy Python scripts, improving throughput by 20%
  • Built hiring demand forecast prototype with MAE of <5%
SQLPythonRegressionForecasting

Technical toolkit

Languages

Python SQL TypeScript JavaScript R Java C++

Cloud & Data

AWS GCP PostgreSQL MongoDB Redis Airflow

Machine Learning & AI

TensorFlow LangChain Scikit-learn PyTorch Keras Transformers Vertex AI

Full-Stack & Tools

React Next.js Tailwind CSS Node.js Recharts Pandas NumPy Docker Git

Selected projects

01

Vazmora — Financial Document Intelligence

AI-powered SaaS that auto-extracts every KPI from annual reports, 10-Ks, and earnings filings — revenue, margins, EPS, EBITDA, P/E ratios — with no templates. Nine tools in one upload: text, tables, OCR, 35+ language translation, editing, chat with page citations, invoice extraction, and multi-document quarterly comparison.

PythonFastAPINext.jsLangChainQdrantCloudflare R2
Live
02

Intelligent Document Processing & Translation

Multi-language translation engine preserving complex layouts, tables and columns. Supports Hindi, Arabic, Cyrillic. Instant extraction of financial documents plus multimodal Q&A capabilities.

PythonStreamlitLLMsOCR
Live
03

VoiceTable — AI Restaurant Operations

Full-stack AI platform reducing RTO on COD orders via automated voice confirmations. Analytics dashboard with confirmation rates, revenue, and sentiment. Real-time webhooks with call transcripts.

Next.js 14TypeScriptPostgreSQLBolna AI
Live
04

CRM Data Extractor Chrome Extension

Chrome Extension for ActiveCampaign extracting Contacts, Deals, and Tasks via adaptive CSS selectors and shadow DOM feedback. React popup dashboard with search, filtering, and CSV/JSON export.

ReactTypeScriptChrome APINode.js
05

Movie Assistant Chatbot RAG Pipeline

Conversational AI using Retrieval-Augmented Generation to query IMDb Top 1000 data. 60% faster retrieval through FAISS vector store and open-source LLMs.

LangChainTransformersFAISSRAG
06

RecursoLM — OpenAI Parameter Golf

Recurrent LLM with 2 shared transformer blocks reused 16x. ~3.15M unique parameters within a 16MB artifact budget. GQA, ALiBi positional bias, SwiGLU, learned recurrence gate, INT8+zlib export. Trained on FineWeb shards.

PyTorchTransformersGQAALiBiSwiGLUDistributed

Education & credentials

MSc Data Science

Christ University, Bengaluru

2025 — 2027

BSc (Hons) Computer Science

Delhi University, New Delhi

2020 — 2023

Reliance Foundation PG Scholarship

Shortlisted for academic excellence

2025

Marketing Head — Expectation'25

Record-breaking fest participation

2025

Stanford — Machine Learning

Coursera certification

Cert

Data Science using Python

University of Delhi

Cert

Let's work together

Open for full-time roles, freelance projects, or just a good conversation about AI and data. Let's connect.

avinashvs0052@gmail.com
+91 7052985015 Bengaluru, India