Pavan Sekhar Mandavilli profile photo 1 on Tal

Pavan Sekhar Mandavilli

Serving Notice Period | AI/ML Engineer | Agentic AI & LLM Systems | Multi-Agent Systems | MCP | LangGraph | RAG | Python · FastAPI · GCP

Bengaluru

4+ years of work experience

Kapture CX
Nov 202511 mosPresent

Bengaluru

Building production-grade Agentic AI and LLM platforms at Kapture CX, focusing on AI tool orchestration, multi-agent systems, MCP infrastructure, RAG, and scalable AI services. • Designed and deployed a multi-tenant MCP infrastructure on GCP for secure AI tool registration, isolation, and live discovery across MCP-compatible clients, implementing OAuth2/JWT-based access control. • Built a LangGraph-based multi-agent orchestration platform with serverless graph execution, state checkpointing, and real-time streaming for fault-tolerant and resumable AI workflows. • Developed a visual agent workflow builder enabling non-technical users to configure agents, connect tools, and define inter-agent workflows without writing code. • Built and deployed FastAPI-based AI orchestration services supporting 1,000+ daily requests with 90% uptime across distributed workflows. • Built an intelligent document-processing and RAG pipeline using embeddings and semantic search, improving retrieval accuracy by 35%. • Implemented Redis-based inference caching, reducing API latency by 20% and increasing throughput by 35%. • Designed REST APIs and MongoDB/PostgreSQL data models with indexing strategies for scalable AI applications and vector similarity search.
Kognitiv Club
Aug 20242 yrs 2 mosPresent

As an Advisor, I am committed to supporting members in their technical development and creating a collaborative atmosphere. My goal is to assist members in overcoming challenges, exploring innovative ideas, and reaching their project objectives while promoting leadership and teamwork within the club.
Kapture CX
Dec 202411 mosOct 2025

Benguluru

Focused on retrieval systems, ML inference infrastructure, and backend services. • Built an intelligent document-processing pipeline using vector embeddings and semantic search, improving retrieval accuracy by 35% through optimized pipeline design and feature engineering. • Implemented Redis caching for ML inference results, reducing API latency by 20% and improving throughput by 35%. • Designed RESTful APIs and database schemas across MongoDB and PostgreSQL, with indexing strategies optimized for vector similarity search at scale. • Contributed to production RAG pipelines powering LLM features across the platform.
Kognitiv Club
Aug 20231 yr 3 mosOct 2024

As the Director of Technology at Kognitiv Technology Club, I contribute my skills by leading projects and managing technical events. I oversee the execution of innovative initiatives and ensure seamless event coordination. My role includes guiding the team to achieve technical excellence and fostering a collaborative environment. Through strategic planning and hands-on involvement, I drive the club's technological growth.
Kognitiv Club
Jul 20221 yr 2 mosAug 2023

As an active member of the Kognitiv Technology Club, I participate enthusiastically in club events. I assist senior members in executing events successfully and contribute to the club's initiatives. My involvement supports the club's overall growth and success.
GitHub
@pavanmandavilli
Personal portfoliopavanmandavilli.vercel.app

Pavan’s AI Stack

elevenlabs
elevenlabs
hugging face
hugging face
langchain
langchain
langgraph
langgraph
mcp
mcp
Browser Use
Browser Use

Education

kl university

Bachelor of Technology - BTech, AI DS

kl university

20214 yrs2025

Join Pavan and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Pavan Sekhar Mandavilli on Tal

Headline

Serving Notice Period | AI/ML Engineer | Agentic AI & LLM Systems | Multi-Agent Systems | MCP | LangGraph | RAG | Python · FastAPI · GCP

About

AI/ML Engineer with nearly 2 years of professional experience building production-grade Agentic AI, LLM systems, and AI infrastructure. My main focus is on Agentic AI, LLM applications, multi-agent systems, MCP, and RAG. I enjoy working on the engineering side of AI — taking an idea or prototype and turning it into something that can actually run reliably in production. At Kapture CX, I’ve been working on problems around AI agent orchestration, tool integration, retrieval, and scalable AI services. Some of the systems I’ve built include a multi-tenant MCP platform, a LangGraph-based multi-agent orchestration platform, and RAG-based document processing pipelines. A few results from my work: • Reduced API latency by 20% using Redis-based inference caching • Improved retrieval accuracy by 35% through RAG and data-processing improvements • Increased system throughput by 35% • Built AI services handling 1,000+ requests per day I work mainly with Python, LangGraph, LangChain, MCP, FastAPI, GCP, Redis, PostgreSQL, vector databases, and Docker. What interests me most is figuring out how to make AI systems more reliable, scalable, secure, and useful in real-world applications. I’m currently looking to connect with people working on Agentic AI, LLM infrastructure, AI platforms, and GenAI products.

Location

Bengaluru

AI stack

elevenlabs

hugging face

langchain

langgraph

mcp

Browser Use

Work experience

Software Engineer – AI/ML at Kapture CX, Nov 2025 - Present · 11 mos, Bengaluru, Building production-grade Agentic AI and LLM platforms at Kapture CX, focusing on AI tool orchestration, multi-agent systems, MCP infrastructure, RAG, and scalable AI services. • Designed and deployed a multi-tenant MCP infrastructure on GCP for secure AI tool registration, isolation, and live discovery across MCP-compatible clients, implementing OAuth2/JWT-based access control. • Built a LangGraph-based multi-agent orchestration platform with serverless graph execution, state checkpointing, and real-time streaming for fault-tolerant and resumable AI workflows. • Developed a visual agent workflow builder enabling non-technical users to configure agents, connect tools, and define inter-agent workflows without writing code. • Built and deployed FastAPI-based AI orchestration services supporting 1,000+ daily requests with 90% uptime across distributed workflows. • Built an intelligent document-processing and RAG pipeline using embeddings and semantic search, improving retrieval accuracy by 35%. • Implemented Redis-based inference caching, reducing API latency by 20% and increasing throughput by 35%. • Designed REST APIs and MongoDB/PostgreSQL data models with indexing strategies for scalable AI applications and vector similarity search.

Advisor at Kognitiv club at Kognitiv Club, Aug 2024 - Present · 2 yrs 2 mos, As an Advisor, I am committed to supporting members in their technical development and creating a collaborative atmosphere. My goal is to assist members in overcoming challenges, exploring innovative ideas, and reaching their project objectives while promoting leadership and teamwork within the club.

Software Developer Trainee at Kapture CX, Dec 2024 - Oct 2025 · 11 mos, Benguluru, Focused on retrieval systems, ML inference infrastructure, and backend services. • Built an intelligent document-processing pipeline using vector embeddings and semantic search, improving retrieval accuracy by 35% through optimized pipeline design and feature engineering. • Implemented Redis caching for ML inference results, reducing API latency by 20% and improving throughput by 35%. • Designed RESTful APIs and database schemas across MongoDB and PostgreSQL, with indexing strategies optimized for vector similarity search at scale. • Contributed to production RAG pipelines powering LLM features across the platform.

Director of Technology at Kognitiv Club, Aug 2023 - Oct 2024 · 1 yr 3 mos, As the Director of Technology at Kognitiv Technology Club, I contribute my skills by leading projects and managing technical events. I oversee the execution of innovative initiatives and ensure seamless event coordination. My role includes guiding the team to achieve technical excellence and fostering a collaborative environment. Through strategic planning and hands-on involvement, I drive the club's technological growth.

Club Member at Kognitiv Club, Jul 2022 - Aug 2023 · 1 yr 2 mos, As an active member of the Kognitiv Technology Club, I participate enthusiastically in club events. I assist senior members in executing events successfully and contribute to the club's initiatives. My involvement supports the club's overall growth and success.

Education

Bachelor of Technology - BTech, AI DS, kl university

Public links

github: https://github.com/pavanmandavilli

pavanmandavilli.vercel.app: https://pavanmandavilli.vercel.app/