by Shloka Kulkarni

I BUILD AI SYSTEMS
THAT WORK IN PRODUCTION.

—SHLOKA KULKARNI · AI ENGINEER · TECHNOSSUS AI STUDIO

RAG Pipelines · Voice Bots · LLM Engineering · Intelligent Document Systems

Shloka Kulkarni is an AI Engineer based in Pune, India, specialising in Retrieval Augmented Generation, LLM engineering, Python backend development, FastAPI, pgvector, voice bots, and intelligent document processing. Currently at Technossus AI Studio. Previously at CrowdStrike.

Experience

AI Engineer · Technossus — AI Studio
Jan 2026 – Present  ·  Pune
  • —Researching, experimenting and shipping AI-first products as part of a dedicated AI Studio team.
  • —Built production RAG systems, intelligent document pipelines, voice bots, and agentic tooling.
  • —End-to-end ownership from architecture through deployment for AI-native features.
Backend Developer · Kashnate Solutions
Jul – Dec 2025  ·  Pune
  • —Built Python backend services and AI-assisted Node.js frontend features for client-facing web apps.
  • —Integrated LLM-powered features to accelerate content generation and workflow automation.
Malware Research Engineer · CrowdStrike
Jun 2024 – Jun 2025  ·  Remote
  • —Deployed production services in Python, TypeScript, and React across distributed security infrastructure.
  • —Built a fault-tolerant scheduling microservice that reduced processing delays by 30%.
  • —Worked across Docker, Kubernetes, Rancher, OpenSearch, and S3.
Backend Software Engineer · Comarete Technologies
Jan – Jun 2024  ·  Pune
  • —Built Python microservices and REST APIs for production client applications.
  • —Achieved a 20% performance improvement through targeted query optimisation.
  • —Stack: pytest, MySQL, MongoDB.

Projects

LiveMind

6 techs

An AI chatbot you can deploy on any website — it scrapes the site live, learns it instantly, and answers user questions conversationally. No document uploads, no storage setup.

PythonFastAPIpgvectorPlaywright+2 more

Knowledge Engine

4 techs

Enterprise RAG backend — chunker → embedder → vector store → re-ranker → LLM response. Full document lifecycle management and test suite.

PythonFastAPIpgvectorRe-ranker

SentiCore

5 techs

AI-powered patient appointment system over WhatsApp — no app, no website. Patients chat with an AI that takes symptoms, suggests the right specialist, and manages queues.

AIWhatsApp IntegrationNLPPython+1 more

OnBoardIQ

4 techs

AI Background Verification pipeline — classifies KYC docs, deduplicates across 3 layers, organises per-candidate, generates audit reports.

PythonOllamaTesseract OCRStreamlit

Smart Revenue Collector

5 techs

Full-stack AI debt-collection system: ML defaulter scoring, LLM outreach generation, priority queue, live React dashboard.

PythonFastAPIReactTypeScript+1 more

VoiceBot PoC

4 techs

Multi-service voice assistant — Python STT/TTS microservice + C#/.NET orchestrator with Clean Architecture.

PythonFastAPIC#.NET

BillSage AI

4 techs

Async OCR pipeline for intelligent bill classification with analytics dashboard.

PythonOCRFastAPIStreamlit

Technossus Design System

Live ↗
6 techs

Figma-to-code design system with tokens, Tailwind preset, and React components. Built full website pages from the design system alone.

TypeScriptReactTailwind CSSFigma+2 more

Skills

AI & ML
LLMsRAGpgvectorOllamaOCRNLPscikit-learnCNNs
Backend
PythonFastAPIC#/.NETNode.jsREST APIsWebSockets
Frontend
ReactTypeScriptNext.jsTailwind CSSVite
Databases
PostgreSQLMongoDBMySQLOpenSearchElasticsearchRedisS3
DevOps
DockerKubernetesRancherAzure DevOpsGitHub ActionsAWS

Contact

LET'S BUILD
SOMETHING.

Open to full-time roles, freelance projects, and interesting conversations.

psst... ask me something 👀
pixel cat mascot