About Vishwa Teja Vangari

I started as a Java developer in Hyderabad in 2011, building dealer communication and sales systems for Nissan Europe at Tech Mahindra. JSF, EJB, Hibernate over Oracle, WebSphere and WebLogic, release trains through Jenkins. That job taught me how enterprise software works. Many people depend on a system they never see, and the release process exists because someone once broke production.

In 2014 I moved to New Orleans for a master's in computer science at the University of New Orleans, after a B.Tech. in information technology at Kakatiya Institute of Technology and Science in Warangal. I worked as a graduate assistant building web and iPad applications for research projects, and graduated in December 2015.

I joined Egen in February 2016 as a software developer and have been there since. The first few years were healthcare: clinical data into the OMOP common data model on Azure, Spring Boot and Scala services, Angular and React front ends. From September 2018, as a staff engineer, a long stretch of streaming work. Kafka into Redshift, Elasticsearch and PostgreSQL, a rules engine on Kafka Streams, Airflow and PySpark pipelines, single sign-on with Ping Identity. I wrote about the Kafka part on Medium in 2019 and published the tooling on GitHub. From March 2021, as a principal engineer, I ran solution architecture and design reviews across projects and mentored engineers on their service designs.

Since April 2024 I've been Senior Director of Engineering. I lead the application, data and AI engineering teams for a large retail client, and the team behind the platform our teams use to build and run AI agents. I lead the architecture for the order and fulfillment services and their Azure cloud digitization, the AI agents built around them, and the governed data platform they read from, and I work with the engineering teams that deliver them. On the platform side I set the shape of the agent runtime, the ontology and semantic layer under the agents, and the evaluation suites each release must pass, and the platform team builds and runs them. I also led the design of Egen's agentic engineering harness; the platform team built it, and our delivery teams moved onto the agentic SDLC that runs on it.

The common thread is the order lifecycle: orders, inventory, picking, substitutions and delivery. Most of what I know about platforms came from systems where a wrong decision shows up as a missing item in a customer's order.

On my own time I study ontology-grounded agents, for my own learning. The question I keep working on is what has to be written down, and by whom, before an agent can be trusted with a business question. I build small practice experiments on synthetic data in healthcare, retail and facilities operations. It is for learning only. It is not work for any company or client. The notes are on the learning page and the engineering harness page, and I write about what I learn.

Outside the client work, I write the AI, data and applications parts of Egen's strategy with the leadership team, run the hiring loop for software, AI and data engineering roles, and run the technical workshops and bootcamps for our teams. I'm an editor of the Egen Engineering & Beyond publication on Medium.

Speaker bio

Vishwa Teja Vangari is Senior Director of Engineering at Egen, where he leads the application, data and AI engineering teams for a large retail client and the team behind the platform those teams use to build and run AI agents. He led the architecture of the platform's agent runtime and the ontology, knowledge graph, semantic layer and context layer under it, and the design of the agentic engineering harness that Egen's teams deliver with. On his own time he studies ontology-grounded agents and writes about what he learns. His background is in order management and fulfillment systems, streaming data platforms, and clinical data in healthcare. He holds an M.S. in Computer Science from the University of New Orleans and is an AWS Certified Solutions Architect, Professional, and a Microsoft Certified Azure Solutions Architect Expert. He lives in Naperville, Illinois, in the Greater Chicago Area.

Short bio

Vishwa Teja Vangari is Senior Director of Engineering at Egen, where he leads the teams that build agent platforms, data platforms and order systems for a large retailer. He studies ontology-grounded agents on his own time and writes about what he learns at vishwatejavangari.com.

Speaking

I'm open to speaking at meetups, conferences and internal engineering events. Each talk runs 25 or 45 minutes and can be pitched to engineers or to a leadership audience. Send me a message on LinkedIn with the event, the audience and the date.

Agents in production: the loop, the permissions and the gates

What an enterprise needs in place before it lets an agent act. The runtime loop, tool scope, human review, evaluation suites, and the order to build them in. Drawn from a platform that runs agents for fulfillment, product catalog work and marketing.

Similarity is not knowledge

One audit question, a synthetic cohort with planted answers, and two architectures. Where retrieval loses the answer, what an ontology finds, and what it costs in tokens and time. The audience leaves with a method they can apply to their own questions.

The engineering harness: shared agents, blocking checks, and what engineers do now

Coding agents are moving from one engineer's laptop to something the whole organization shares. What that takes: shared context, permissions, checks that block, and the record a release leaves behind. Also what changes for hiring, review and KPIs.

One data platform for dashboards and agents

Serving analytics and agents from one governed copy. Mirroring as the ingestion backbone, semantic models scoped by business domain, a freshness target measured in seconds, and the cost model that goes with it.

Retail order fulfillment

Picking online orders inside stores. Labor-aware picking, dedicated pick areas replenished on their own, and substitutions that customers accept.