Looking for a job

Johin Johny

@johinjohny

Message

I build production AI platforms, multimodal search engines, and automated market intelligence systems.

India

Message

What I'm looking for

I'm looking to build impactful AI-driven products at the intersection of machine learning and software engineering, especially intelligent systems that improve search, retrieval, automation, and business decision-making.

At Glassbox, I built and deployed a production market-intelligence platform used daily by the business-development team.

I engineered an automated pipeline that classified 40,000+ market signals, matched them to clients and competitors, and drafted outreach emails. I also designed an 11-stage AI agent lifecycle for intelligence briefs, pitch decks, and proposals with review gates and audit trails.

I launched a multimodal fashion search engine that ingested 10K+ Westside products and supported image- and text-based retrieval. Using FashionCLIP, SigLIP, and Qdrant, I delivered sub-100 ms query latency and selected SigLIP through retrieval benchmarking.

Previously at Augle AI, I built a RAG chatbot, knowledge graph, and conversational frontend, improving answer accuracy by 30% and reducing graph traversal time by 50%.

Experience

Work history, roles, and key accomplishments

Artificial Intelligence Engineer

Glassbox

Jan 2026 - Aug 2026 (7 months)

Built and deployed a production market-intelligence platform (Next.js 16, TypeScript, Neon Postgres, OpenAI API) used daily by the business-development team, live at gbhunt.vercel.app. Engineered an automated pipeline that scrapes and LLM-classifies market signals by category and priority, matches each to tracked clients or competitors, and auto-drafts outreach emails; scaled to 40,000+ classified

AI Developer

Glassbox

Apr 2025 - Jan 2026 (9 months)

Architected and launched a multimodal fashion search engine, scraping and ingesting 10K+ Westside products to power integrated image- and text-based retrieval.
Generated joint embeddings with FashionCLIP & SigLIP, storing vectors in Qdrant to deliver sub-100 ms query latency.
Benchmarked OpenAI CLIP vs. FashionCLIP vs. SigLIP, selecting SigLIP after it achieved the highest top-1 retrieval accuracy

Software Developer Intern

Augle AI

Apr 2024 - Apr 2025 (1 year)

Developed a RAG chatbot using LangChain, PromptLayer, and the OpenAI API, improving answer accuracy by 30%.
Scraped and normalized 5,000+ product records with Scrapy, creating a clean dataset for knowledge-graph and RAG ingestion.
Modeled and populated a Neo4j knowledge graph to accelerate semantic retrieval, reducing node-traversal times by 50%.
Connected a React.js frontend to Flask REST endpoin