# Krish Parekh — AI Engineer in Melbourne, Australia

Source: https://krishparekhdev.com/

I build RAG systems, multi-agent applications and voice agents, along with the APIs and cloud infrastructure behind them. Right now that means production voice agents and ad-creative pipelines for Australian businesses.

My work spans Python, FastAPI, Amazon Bedrock and Terraform, alongside Next.js and TypeScript. I care about making complex systems simple to use, and I share what I learn through open-source projects and interactive technical articles.

## Experience

### AI Engineer — SMEC AI

Melbourne, Australia · 03.2026–Present

- Built the production voice agent for SMEC AI's AI Information Line to handle concurrent, multi-turn calls and answer enquiries before callers book a consultation. This reduced unqualified discovery calls by 50% and eased the team's workload. The team also observed an increase in phone consultations as callers became more comfortable using the service.
- Built an ad-creative pipeline connecting fal, ElevenLabs, OpenAI and HyperFrames. Campaigns using these creatives reduced cost per lead by about 93–94% and generated 200K+ impressions across multiple channels.
- [National AI Centre: Got questions about AI? Now there's a number you can call.](https://www.ai.gov.au/news-and-insights/blog/got-questions-about-ai-now-theres-number-you-can-call)
- [MFAA: New Australian AI Information Line launched: A practical resource for brokers and small businesses](https://www.mfaa.com.au/news/new-australian-ai-information-line-launched-a-practical-resource-for-brokers-and-small-businesses)
- [business.gov.au: SMEC AI](https://business.gov.au/expertise-and-advice/smec-ai) (service listing)
- [ACS Information Age: New AI advice line for business is run by… AI](https://ia.acs.org.au/article/2026/new-ai-advice-line-for-business-is-run-by--ai.html)
- [iTWire: SMEC AI launches the AI Information Line: free national AI help for small business, in one phone call](https://itwire.com/business-it-news/data/smec-ai-launches-the-ai-information-line-free-national-ai-help-for-small-business-in-one-phone-call)
- Skills: Python, FastAPI, Next.js, TypeScript, LangChain, n8n, AWS

### Software Engineer Intern — Capyra

Melbourne, Australia · 07.2025–11.2025

- Migrated the app from Ionic to Next.js, improving page load performance by 40% with SSR and raising the Lighthouse SEO score from 0 to 96.
- Rebuilt backend services from JavaScript to Python (FastAPI) with WebSocket streaming for real-time STT/TTS, cutting P95 latency from 1s to 450ms and lifting translation accuracy 18% via Google Vertex AI.
- Optimised container builds with multi-stage Docker, cutting the frontend image from 1.2GB to 400MB and the backend from 743MB to 537MB, with automated GCP Artifact Registry and Cloud Run deploys.
- Skills: Next.js, TypeScript, Python, FastAPI, Docker, GCP

### Software Engineer — Ethan.AI

Bengaluru, India · 01.2023–02.2024

- Built a multi-agent financial analysis system over bank statements: LlamaParse OCR into structured transactions, LlamaIndex retrieval over a multi-tenant Qdrant vector database, and a ReAct agent with dynamic tool calling across third-party APIs.
- Migrated the frontend from JavaScript/React to Next.js/TypeScript with Ant Design, cutting code size 55%, improving SEO via SSG, lowering LCP/FCP, and reducing UI/UX defects 40%.
- Optimised report-generation APIs with improved SQL, caching layers, and cron pipelines, improving P95 API performance by 30%.
- Automated infrastructure with Terraform and CI/CD, migrating the backend from manual AWS to GCP and the frontend to AWS Amplify, reducing deployment costs by 30%.
- Built a database crawler for automated bug detection, saving an hour of operational time daily and lifting data quality from 75% to 98%.
- Skills: Python, LlamaIndex, Qdrant, Next.js, TypeScript, PostgreSQL, Terraform, AWS

### Android Developer Intern — Brainly Panda

Remote · 05.2022–08.2022

- Upgraded the food-delivery fare module with the Google Maps Matrix API for dynamic pricing on distance, traffic, and fastest route, increasing revenue 65%.
- Built a full-stack Android data entry app with an Express.js backend, replacing manual processes and improving operational efficiency by 40%.
- Skills: Kotlin, Android, Express.js, Firebase

## Education

- Master of Information Technology, RMIT University (02.2024–12.2025)
- Bachelor of Science, Usha Pravin Gandhi College of Arts, Commerce and Science (2020–2023)

## Contact

- [Email Krish Parekh](mailto:krishparekh303@gmail.com)
- [Resume](https://krishparekhdev.com/Krish%20Parekh%20Resume.pdf)