About Experience Skills Projects Contact

AI Engineer & Tech Lead — Sydney, AU

MahitChoubey

_

I architect agentic AI systems that work in production — not just in demos. From multi-agent pipelines and LLM infrastructure to the teams and technical decisions that ship them at scale.

View My Work Get In Touch
2+ yrs
AI Engineering
10+
Engineers Led
Scroll

Building AI at the
edge of what's possible

I'm an AI Engineer and Tech Lead based in Sydney, Australia. I build intelligent systems end-to-end — from agent architecture and LLM orchestration to the products and teams that surround them.

At Apate.AI I lead a cross-functional team of 10+ onshore and offshore engineers working on some of the most technically demanding agentic AI applications in the security intelligence space. I've been heavily involved in the development and architecture across the core products — from real-time conversational AI systems to autonomous OSINT agents operating across live platforms.

What drives me is the gap between AI demos and AI that actually works in production. The architecture decisions, the failure modes, the engineering discipline — that's where I operate. I'm equally comfortable whiteboarding a system design and writing the code that ships it.

const mahit = {
  role: "Tech Lead",
  company: "Apate.AI",
  location: "Sydney, AU",
  focus: [
    "Agentic AI Systems",
    "LLM Architecture",
    "Engineering Leadership",
  ],
  clouds: ["AWS", "Azure", "GCP"],
  building: "always",
};

Where I've built things

Apate.AI
Apr 2025 – Present
Tech Lead
  • Lead architect and team lead across a cross-functional team of 10+ onshore and offshore engineers, managing delivery, architecture decisions, and technical direction as the company scaled through seed funding and expanded across new markets and product lines
  • Heavily involved in the design and build of the foundational AI pipeline infrastructure powering Apate's real-time disruption systems — deployed across Tier-1 Australian telcos and Big 4 financial institutions
  • Engineered AI-powered scam baiting agents deployed at scale across voice and SMS channels, engaging thousands of active scammers daily and extracting live intelligence
  • Built autonomous OSINT AI agents that infiltrate platforms like Telegram to discover and map crypto scam rings and banking fraud networks — extracting wallet addresses, phone numbers, and operational playbooks from live criminal operations
  • Drove end-to-end technical decisions across agentic AI systems, real-time data pipelines, and cross-sector intelligence sharing infrastructure
Agentic AILangChainOSINTAWSGCPSystem DesignTeam LeadLLMOpsLLM Evals
Apate.AI
Nov 2024 – Apr 2025
Software Engineer
  • Joined as a founding engineer at pre-seed stage — part of the core team that took Apate from early product to seed-funded commercial deployment across Tier-1 Australian telcos and major financial institutions.
  • Contributed to core AI product development across voice and text-based conversational agent systems, building real-time conversation management infrastructure
  • Developed intelligent data extraction and enrichment pipelines feeding structured intelligence to banking and telco partners
  • Worked full-stack across the web platform, backend services, and cloud deployment on Azure
PythonReactNode.jsLLMsAzureTypeScriptVoice AIConversational AI
Stickman Cyber
Sep 2023 – May 2024
Junior AI/ML Developer
  • Heavily involved in the testing and implementation of fine-tuned open-source LLMs applied to cybersecurity policy compliance assessments — including ISO framework evaluations
  • Worked across the implementation lifecycle from model evaluation through to production integration, contributing to a pattern of bridging AI research and real-world software delivery
PythonMachine LearningNLPHuggingFaceTypeScriptCybersecurity
Macquarie University
2021 – 2024
B.Eng Software Engineering
  • Specialised in systems design, algorithms, and applied AI — the technical foundation for a career in production AI engineering

Tech I work with

// AI Engineering
LangChainLangGraphLlamaIndexCrewAIAutoGenOllamaRAGMulti-Agent SystemsMCPClaudeOpenAIGeminiHuggingFaceVector SearchLLM EvalsFine-tuningElevenLabsDeepgramVoice AI
// Cloud & Infrastructure
AWSAzureGCPAzure OpenAIAWS BedrockVertex AIDockerKubernetesCI/CDServerlessRedisPostgreSQLDynamoDBClickHouseAirbyteApache SupersetMetabase
// Software Engineering
PythonTypeScriptNode.jsReactNext.jsREST APIsWebSocketsGit
// Specialisations
Agentic System DesignAgent OrchestrationEvent-Driven AI ArchitectureConversational AIVoice AgentsOSINT & Threat IntelligenceLLMOpsEngineering Leadership

Things I'm building

01 / Education Tech · Deployed
Realtime AI Feedback for Cybersecurity Education
A lightweight Chromium browser extension that embeds an AI copilot directly into cybersecurity learning modules — delivering real-time, contextually-aware feedback as students work through technical exercises. Designed for easy integration into existing learning platforms with no infrastructure changes required on the institution's side. Reduces the feedback loop in technical education from days to seconds, enabling students to course-correct in the moment rather than after the fact.
02 / Productivity · In Development WIP
WFH Chief of Staff
An ambient AI agent for remote workers — the moment you step away from your desk it becomes an active observer across your key work stack: Slack, Outlook, Jira, and any running coding agents. When you return, it surfaces a prioritised intelligence brief: critical updates, pending reviews, and decisions that can't wait — with full context, not just notifications. Built around multi-agent orchestration and MCP for deep, real-time integration across tools.

Let's connect

Let's build something worth building.

Whether it's a collaboration, a technical discussion, or an interesting problem in the AI space — I'm always up for a good conversation.