Senior AI Engineer Bengaluru, India

Portfolio 2026

Vishnu Prasad

I build production AI and backend platforms around agentic systems, retrieval, evaluation, and the unglamorous work that keeps them reliable.

Explore selected work Start a conversation
Experience
9+ years
Platform uptime
99.9% delivered
Enterprise reach
50+ customers
Manual investigations
30% fewer
01 Selected impacts

01 Selected impacts

Work measured in production outcomes.

A few concrete results from AI and platform work across enterprise security, qualitative research, and developer tooling.

  1. 01 / Security AI

    Less manual investigation

    Built multi-agent AI automation for log analysis, case prioritization, and threat hunting at Trend Micro.

    30%fewer manual security investigations

  2. 02 / Platform delivery

    Reliable enterprise operations

    Led Python and Go backend delivery for Vision One SOAR and MXDR platforms serving enterprise customers.

    99.9%uptime across 50+ enterprise customers

  3. 03 / Retrieval

    Evidence-led model selection

    Developed a RAG pipeline over security documents and an evaluation framework comparing OpenAI, Anthropic, and open-source models on accuracy, latency, and cost.

    1,000+security documents in retrieval

  4. 04 / Developer tools

    Multi-agent debugging in practice

    Created a Python package for multi-agent log analysis and API-failure debugging across hosted and local model providers.

    500+debugging sessions

02 Working approach

Production AI is a system, not a prompt.

My work connects data foundations, retrieval, orchestration, evaluation, and observability into one maintainable path.

A practical pipeline for AI product delivery
  1. 01 Foundation

    Data & services

    Python, Django, FastAPI, PostgreSQL, Redis, S3, Kafka, and event-driven workers.

  2. 02 Context

    Retrieval

    Embeddings, semantic search, document pipelines, and RAG over operational knowledge.

  3. 03 Orchestration

    Agents

    Multi-agent workflows with LangGraph, CrewAI, LangChain, AutoGen, and OpenAI Agents SDK.

  4. 04 Evidence

    Evaluation

    Model comparisons and quality checks across accuracy, latency, cost, and delivered behavior.

  5. 05 Operations

    Observability

    AWS EKS, Kubernetes, CI/CD, OpenTelemetry, and production practices built for uptime.

03 Selected work

Systems with a reason to exist.

Open-source tools and product work spanning multi-agent debugging, orchestration, retrieval, and developer productivity.

Developer tooling Python package

Multi-Agent Debugger

Multi-agent log analysis and API-failure debugging with support for OpenAI, Anthropic, Google, and Ollama. Used across 500+ debugging sessions.

View on GitHub
Agent systems Orchestration

GemMate

An agent-team orchestration platform with search, file analysis, memory, and an LLM evaluation dashboard.

View on GitHub
Agent systems Framework + CLI

CrewAI Master

A framework and CLI for generating and managing autonomous agents with memory and RAG context retrieval.

View on GitHub
Open source Developer utilities

j2toon / mask-env

Python utilities for JSON-to-TOON conversion, AI tooling, and developer productivity workflows.

Explore the repositories

04 Career

From backend systems to agentic platforms.

A decade of building, leading, and operating software across enterprise platforms, security, data, and AI.

  1. Jul 2026 – Present

    Scopely

    Senior Agentic AI Engineer

    Working in the central GenAI team to enable AI adoption and build applications with coding agents. Enabled company-wide AI observability with OpenTelemetry, co-developed multi-agent delivery workflows, and guided implementation decisions and code quality.

  2. Dec 2025 – Apr 2026

    HeyMarvin

    Senior Backend Developer

    Owned a backend data-ingestion layer for an AI-powered qualitative research platform, including Django services, FastAPI workers, retrieval pipelines, PII detection and redaction, audit logging, and test coverage.

  3. May 2023 – Nov 2025

    Trend Micro

    Senior Technical Lead / Senior Software Engineer

    Led Python and Go delivery for enterprise security platforms, built multi-agent AI automation, and developed RAG and model-evaluation systems. Architected AWS EKS microservices, delivered service integrations, and mentored a 10-member engineering team.

  4. Jul 2018 – Apr 2023

    Knowledge Lens

    Senior Technology Lead / Technology Lead

    Owned architecture and delivery for enterprise Python platforms. Built IoT monitoring for 5,000+ devices, license-management and self-healing systems, and REST and AWS Lambda/S3 analytics pipelines.

  5. Oct 2016 – May 2018

    Tech Mahindra

    Associate Software Engineer

    Automated secure file exchanges, Linux health checks, and memory diagnostics with Python and shell scripting, while handling 24x7 production incident response and root-cause analysis.

05 Capabilities

Deep in the stack. Fluent in the system.

AI engineering grounded in backend fundamentals, distributed systems, and the operational details that make products dependable.

01

AI & LLM engineering

Agentic AI, multi-agent systems, RAG, LLM integration, prompt engineering, LLM evaluation, embeddings, semantic search, LangGraph, CrewAI, LangChain, AutoGen, OpenAI Agents SDK.

02

Languages

Python, Go, Rust, JavaScript, TypeScript.

03

Backend & architecture

FastAPI, Django, Flask, Celery, SQLAlchemy, Pydantic, REST, GraphQL, microservices, event-driven architecture, distributed systems.

04

Cloud & reliability

AWS EKS, Lambda, S3, SQS, RDS, API Gateway, GCP, Azure, Docker, Kubernetes, Terraform, CI/CD, PostgreSQL, Redis, MongoDB, Kafka, RabbitMQ, Pytest, Prometheus, Grafana, OpenTelemetry.

07 Resume

The complete professional record.

Experience, selected projects, technical scope, education, and certifications in one PDF.

Download resume Open PDF
First page of Vishnu Prasad's resume, showing his senior AI engineer profile, experience, and technical skills.

08 Contact

Building AI that has to work in production?

For conversations about production AI, agentic systems, retrieval, platform reliability, and measurable outcomes, get in touch.

vishnuprasadapp@gmail.com LinkedIn GitHub +91 98804 14861