Headline

founding engineer @ Arctan | DTU 25

About

I’m Aaryan, software engineer based in Bengaluru. At Arctan, I work on Voice AI systems. That has meant building a Rust SDK for audio models inference, writing the Python and Go services behind them, and shipping software that runs those models directly on a user’s device, yes models running on edge without hiccups! One of the products I built is Eigen, a voice-isolation system that runs in under 10 milliseconds, processes more than 100,000 minutes of audio every day, and reduced speech-recognition errors by 57%. It’s also officially integrated with Pipecat. Before Arctan, I built lending software and deployment systems at HealthCRED. Outside work, I contribute to open source most recently Polar and build side projects whenever an idea feels interesting enough to take from a blank page to something usable. I enjoy early stage problems and paving my way from 0 to 1 where the answer isn’t obvious yet, understand what’s broken, build the first version, put it in front of people, and keep improving it until it works. I like building software that is delightful and make my users happy. I believe MVP is bare minimum and actually a starting point and not the finish line for shipping beautiful software. Always happy to meet people building useful and ambitious things.

Location

Bengaluru

AI stack

codex

opencode

Firecrawl

OpenRouter

Wispr Flow

Zed

Work experience

Founding Engineer at Arctan, Jul 2025 - Present · 1 yr 2 mos, Bengaluru, Karnataka, India, Engineered systems for real time voice AI, spanning backend infrastructure, native audio processing, and on device inference. - Built microservices in Python and Go to support telemetry, authentication, and application and SDK distribution. - Engineered an audio graph library in Rust for composing real time audio processing and inference pipelines. - Developed a desktop application using Tauri that runs ONNX models locally for speech enhancement and accent conversion. - Led 0-to-1 exploration and development of voice AI products, including: > Voice Isolation: A preprocessing layer for voice agents that preserves the primary speaker while suppressing other speakers, background noise, and reverberation. > Speech Enhancement: A real time app designed to improve speech clarity and help BPO agents sound more professional and understandable during customer interactions. Learn more @ https://aaryanmishra.com/demo/

Software Development Engineering Intern at HealthCred, Nov 2024 - May 2025 · 6 mo, Delhi, India, Built a Loan Origination System, a "supertool" as called by the finance team. Reduced the credit underwriting time by ~70%, streamlined borrower's journey while also ensuring scalability of custom loan programs.

Renovatio at Stealth, Jun 2024 - Aug 2024 · 3 mos, Built an AI project management tool for interior designers. Validated through customer pilots and two letters of intent. Built context aware generative AI features on top of existing design drafts and templates.

Education

Bachelor of Technology - BTech, Delhi Technological University (Formerly DCE)

BRILLIANTS CONVENT SCHOOL

Public links

github: https://github.com/mshra