SM

AI Solutions Architect · Engineering Leader

Shashi Mahendrakar

I build production multi-agent AI systems and get them working inside real businesses, backed by 15+ years of taking enterprise platforms from pilot to production.

Sammamish, WA (Seattle area) · Open to Forward Deployed Engineer and AI engineering roles

Email me Request my resume
Portrait of Shashi Mahendrakar
  • 15+years shipping enterprise platforms
  • 30M+users reached by systems I built
  • $3.5M+in measurable cost savings
  • 20+engineers led on critical programs

About

Between the customer and the code

I'm a hands-on technical leader who sits between the customer and the code. For 15+ years I've designed and shipped systems at T-Mobile, Cetera Financial Group and H&R Block that reached 30M+ users and saved more than $3.5M in costs.

Since May 2025 I've focused on agentic AI as an independent consultant. I design multi-agent systems, RAG pipelines and LLM-powered workflows with Claude and OpenAI models, and I integrate them with the APIs, webhooks and data that businesses already run on.

What I enjoy most is the forward-deployed part: working directly with executives and engineering teams to turn a promising AI prototype into something secure, reliable and used every day.

Agentic AI
Multi-agent orchestration, RAG, prompt engineering, MCP, LangChain / LangGraph, vector databases
Engineering
Python, TypeScript / Node.js, Java / Spring Boot, React, PostgreSQL, Kafka
Cloud & delivery
AWS, Azure, Kubernetes, Docker, CI/CD with GitHub Actions
Customer delivery
Technical leadership of 5–20 engineers, executive engagement, security and compliance, financial services

How I work

From first conversation to production

  1. 01

    Discover

    Sit with the customer, map the real workflow and agree on what success looks like.

  2. 02

    Prototype

    Build a working agent on real data quickly, so people can react to something concrete.

  3. 03

    Harden

    Add evaluations, guardrails, security and observability before anyone depends on it.

  4. 04

    Ship

    Integrate with existing APIs, data and systems, and roll out safely.

  5. 05

    Operate

    Monitor, measure the impact and keep improving once it's in daily use.

Projects

Recent AI work

Multi-agent research and document assistant

Problem
Researching a topic, reading long documents and answering follow-up questions eats hours of an expert's day.
What I built
A production-grade multi-agent system: a planner breaks each request into steps and hands them to specialist agents for research, document processing and question answering, each with memory and tool use.
  • Claude
  • OpenAI GPT
  • LangGraph
  • Python
  • MCP

Details available on request

Enterprise knowledge assistant (RAG)

Problem
General-purpose chatbots don't know a company's own documents, and they sometimes make up answers.
What I built
Retrieval-augmented generation that searches enterprise knowledge with vector search and passes the most relevant passages to an LLM, grounding every answer in the company's own content.
  • Python
  • Vector DB
  • LangChain
  • PostgreSQL
  • LLM APIs

Details available on request

AI-powered workflow automation

Problem
Many business processes still depend on people copying data between tools and making routine judgment calls.
What I built
n8n workflows that connect LLMs to REST APIs, databases and SaaS platforms. The LLM classifies, summarizes and routes each item, and the workflow completes the process end to end.
  • n8n
  • LLM APIs
  • REST & webhooks
  • PostgreSQL

Details available on request

Experience

Highlights

  1. Independent Technology Consultant

    May 2025 – Present
    • Architected production-grade multi-agent systems on Claude and OpenAI GPT models for research, document processing and question answering
    • Built RAG solutions and n8n workflow automations that connect LLMs to APIs, databases and SaaS platforms

  2. Cetera Financial Group Software Engineering Manager

    2022 – 2025
    • Owned day-to-day operations of the Smartworks Institutions applications and led an onshore and offshore development team
    • Devised and executed the migration of these applications to AWS, saving $2M a year and cutting release cycles by 40%
    • Took advisor onboarding and workflow platforms from pilot to enterprise-scale production, supporting thousands of daily users
    • Designed vendor integration strategies and met regularly with business stakeholders to align on priorities
    • Improved the accuracy of work estimates for management, and led code reviews and system risk and reliability analysis

  3. T-Mobile Senior Technical Architect

    2016 – 2022
    • Led the software development lifecycle for multiple operations support systems (OSS) applications, from roadmap and high- and low-level design to production
    • Directed 20+ engineers through the Sprint–T-Mobile merger, migrating 30M+ subscribers with 99.99% uptime
    • Reported to the executive steering committee and managed relationships with business leaders up to VP level
    • Approved code changes, cleared security scans with Corporate Security, and designed firewall access for external API integrations
    • Hired and mentored the development and production support teams
    • Built Kafka-based event streaming that cut provisioning latency from minutes to seconds

  4. H&R Block Lead Consultant

    2008 – 2016
    • Led the digital transformation of tax filing platforms used by 20M+ customers a year
    • Moved monoliths to Spring Boot microservices, saving $1.5M a year with zero PCI and IRS audit findings

  5. Education

    M.S. Electrical Engineering, West Virginia University · Leadership Program, University of Washington Bothell

Contact

Let's talk

I'm looking for Forward Deployed Engineer and AI engineering roles where I work directly with customers to get AI into production. Email me to start a conversation, or request my resume and I'll send it over.

Email me Request my resume LinkedIn