k
karthikeyan_ts

Karthikeyan TS

@karthikeyan_ts

Production Grade RAG and Agentic Systems Developer

India
Inglés, Tamil
Parte de la información aparece en idioma inglés.
Sobre mí
Hi! I'm Karthikeyan. I build production RAG and agentic AI systems that run on your real documents, databases and workflows that can handle real data. 1. RAG systems: ingestion, retrieval, reranking, citations 2. AI agents & multi-agent workflows 3. Chatbots over your internal docs 4. LLM integration into your app or API 5. Full-stack web applications Try a demo here: (A multi-agent csv file analyzer. Try asking questions with the sample data) https://supplychain-ecru.vercel.app/ Regards, Karthikeyan TS CS @NIT Trichy Portfolio: https://keyashqa.me/... Lee más

Habilidades

k
karthikeyan_ts
Karthikeyan TS
desconectado • 
Tiempo medio de respuesta: 1 hora

Revisa mis servicios

Sitios web y software con IA
I will build production grade agentic and rag applications
DeUSD 75 / proyectoMás información

Porfolio

Experiencia laboral

triSys

AI/ML Developer

triSys • Tiempo completo

May 2026 - Jul 2026 • 2 mos

1. Built a multi-agent orchestrated system with retrieval-augmented generation (RAG), querying structured, unstructured and real-time data through a single interface. 2. Cut retrieval latency by around 85%, from 51 seconds a query down to 6–8 seconds, by consolidating several purpose-specific models into one warm-started model, and added Fast and Deep modes for precision and speed. 3. Cut multi-agent query time by 47%, from 15.96s to 8.51s, by running the SQL and RAG specialists concurrently rather than one after the other. 4. Removed a 30-second cold start on the first query of a session by pre-loading model weights at service boot.