Looking for a job

Nirmal V

@nirmalv

Message

I build agentic AI products, distributed backends, and cloud-native systems.

India

Message

I've built agentic AI products and distributed backends at ContHunt, Infiknit, and Gyournal, from multi-agent research systems to media-canvas runtimes and realtime voice logging.

At ContHunt, I built a Mastra-based research system that searches TikTok, Instagram, and YouTube with video-level citations. It analyzed up to 500 videos per run and supported 2,454 searches across 262 active researchers.

At Nextiva, I shipped Kubernetes-based Python services that unified customer-support conversations and automated Gemini summarization, sentiment, topic tagging, and clustering for live business clients.

Earlier at Sembly, I built topic intelligence, personalized ranking, recommendations, and efficient LLM systems that improved time on site by 30%, content sharing by 20%, and onboarding completion time by 57%.

Experience

Work history, roles, and key accomplishments

Current

Founding Software & AI Engineer

Current

Gyournal

Apr 2026 - Present (5 months)

Built realtime voice logging for iPhone and Apple Watch, and a Go backend for training intelligence. Managed TestFlight with 35 testers and zero crashes.

Swift iOS Watch OS Go Real Time Speech Pipelines Tool calling

Current

Founding Software & AI Engineer

Current

Infiknit

Dec 2025 - Present (9 months)

Developed an agent runtime for a media canvas using the Vercel AI SDK, and shipped an MCP server for external agents.

Vercel AI SDK MCP TypeScript NodeJS Zod

Current

Founding Software & AI Engineer

Current

ContHunt

Dec 2025 - Present (9 months)

Built a multi-agent research system using Mastra, with multimodal analysis at scale and a distributed AI backend on Cloud Run. Achieved significant user engagement and monetization.

Mastra LangGraph Cloud Tasks Cloud Run LiteLLM PostgreSQL Kubernetes Google Cloud Platform Python TypeScript Go Swift

Software Engineer 2

Nextiva

Apr 2024 - Dec 2025 (1 year 8 months)

Built queue-backed Python services for omnichannel AI orchestration and conversation analytics, shipped on Kubernetes to live business clients.

Python Kubernetes Gemini Queue Based Architectures Conversational Analytics Topic Tagging Clustering

AI Trainer, LLM Alignment

Turing

Aug 2024 - Jan 2025 (5 months)

Created RLHF datasets for Google's LLMs, analyzed model failures, and ensured dataset quality through validation and safety screening.

RLHF SFT Preference Learning Datasets Data Quality Schema Validation Deduplication PII Detection Screening

Data Scientist

Sembly

Mar 2021 - Feb 2024 (2 years 11 months)

Developed topic intelligence, personalized ranking, and recommendation systems, increasing time on site by 30% and content-sharing activity by 20%. Reduced content redundancy by 40% with efficient LLM systems.