Founding Engineer · AI Products

Anmol
Gupta

I build AI products end to end: real-time voice agents, LLM evaluation, and event-sourced backends, from first commit to enterprise pilots.

Now at Yolex Labs Google Summer of Code '22 Co-founder · $100K+ revenue

Five years, shipping.

Founding engineer at early-stage startups since 2020, owning products across the full stack, from infra to evals.

26 days

to take Candoor from zero to a deployed product, single-handedly

~680ms

barge-in on real-time voice agents, with sub-second responses

5M+

Shopify stores analyzed by the AI pipeline at StoresData

$100K+

revenue in nine months as co-founder of Voomp

Experience.

Nov 2025 to Present
Remote

Yolex Labs ↗

Founding Technical Generalist

Own two AI products across the full stack: STORM, AI-generated multi-modal work simulations (voice, code, spreadsheets, canvas), and Candoor, anonymous AI voice interviews.

  • Shipped Candoor single-handedly in 26 days: an F#/Akka event-sourced core, real-time voice interviewer, LLM de-identification and theme clustering, and eval harnesses for every AI step.
  • Cut voice-agent latency to sub-second and barge-in to ~680ms on OpenAI Realtime with dual parallel WebSockets and a dual-VAD design; validated it in an adversarial A/B against LiveKit + Silero + Krisp.
  • Built a position-aware LLM evaluation pipeline extending G-Eval, where judge models score role-specific rubrics over transcripts, canvas state and tool traces.
  • Made simulation generation reliable with a deterministic 31-category role classifier and auto-validation of checkable tasks.
  • Engineered verifiable anonymity: reflection audits, AES-256-GCM sealed submissions with crypto-erasure, and 370 automated tests.
  • Rebuilt 100% of production infrastructure solo in under a week (hosting, DNS, TLS, CI/CD, backups, OAuth) with zero data loss.
  • Ran enterprise pilots with banking, energy, and a 150K-employee IT-services firm.
F#TypeScriptNode.jsPostgreSQLEvent sourcingOpenAI Realtime
Feb 2024 to Dec 2025
Remote · Contract

StoresData ↗

Software Engineer
  • Built an AI-powered pipeline analyzing 5M+ Shopify stores, with snapshot versioning of storefront, product and pricing changes over time.
  • Boosted ingestion throughput 25% with event-driven Inngest workflows, running 500+ concurrent scans with backoff.
  • Cut storage costs 50% by migrating blobs to S3 with Gzip compression.
  • Integrated AI agents with Mastra, Ollama and the Vercel AI SDK for querying and summarization.
TypeScriptBunPrismaPostgreSQLInngestMastra
Sep 2023 to Dec 2024
Los Angeles

Sundial

Founding Software Engineer
  • Delivered an AI assistant for solar-installation code comprehension.
  • Achieved 30% faster async processing with a Java Spring Boot + RabbitMQ backend.
  • Deployed Cloud SQL and VPC networking on GCP, boosting data processing efficiency 40%; set up GitLab CI/CD.
Next.jsSpring BootRabbitMQGCP
May to Aug 2022
Remote

Google Summer of Code

Contributor · Mifos Initiative
  • Integrated a centralized IAM system into Apache Fineract's Payment Hub using OpenID Connect, Kong plugins and Keycloak.
  • Built backend services on Kubernetes, Nginx and Spring Boot, reducing incident reports 30%. View project →
Aug 2020 to May 2021
Kolkata

Voomp Technologies

Co-Founder & Software Engineer
  • Built and sold loan management systems to NBFCs, generating $100K+ revenue in 9 months.
  • Built a co-lending feature on Fineract enabling multi-lender loans.
  • Kept 90% client retention through complex sales cycles.

Toolkit.

Languages

  • TypeScript
  • F#
  • Python
  • Java
  • JavaScript / Node.js

AI & Platforms

  • OpenAI Realtime
  • Mastra · Ollama
  • GCP · Kubernetes · Docker
  • Inngest · RabbitMQ
  • Git · GitHub Actions

Frameworks

  • Next.js · React
  • NestJS · Express
  • Spring Boot
  • Django · Flask
  • XState · Commander.js

Data

  • PostgreSQL
  • MySQL
  • MongoDB
  • Redis
  • Event sourcing

Beyond the code.

I grew up in India, spent a semester in South Korea, a summer doing AI research in Taiwan, then moved to the US for my master's. After three years building across LA, Chicago, and Colorado, I moved back.

20+ countries and counting. I ship from wherever I am, work across time zones, and stay at the cutting edge of what's possible with AI. The constant: building things that matter.

Education.

Indiana University Bloomington

MS, Computer Science · 2021 to 2023

SRMIST, Chennai

B.Tech, Information Technology · 2016 to 2020

Sungkyunkwan University, Seoul

Exchange semester, Computer Science · 2018

Publication.

Behavioral Scoring Based on Social Activity and Financial Analytics

Gupta et al. · Springer, 2021 · Inventive Computation and Information Technologies, LNNS vol. 173. Read on Springer →

Say hello

Let's build
something.

Building something ambitious with AI? I'm always up for a conversation.