About Anu Kumar

What the work actually is

The interesting part of agent engineering is not prompt wording. It is what happens when a tool call fails at three in the morning, when a model returns something plausible and wrong, and when an action needs a human to sign off before it commits. That is where most of this work sits: reasoning loops that can be reasoned about, confidence scoring on generated output, approval gates in front of anything irreversible, and retrieval that ties a decision to real company data rather than model memory.

The other half is ordinary systems engineering, and it is not optional. FastAPI and Node services, PostgreSQL, MariaDB and Redis, Docker and Kubernetes, CI/CD pipelines that deploy without ceremony. A model is one component. The system around it is the product, and it is the part that decides whether anyone trusts the output.

Track record

Currently Senior AI Engineer at SuperManager AGI (Mar 2026 - Present), on the agent and LLM layer of an enterprise agentic platform. Before that, Aug 2023 - Feb 2026 at Mind2Web building full-stack products on React, Node.js and PostgreSQL, including an AI-assisted interaction platform on FastAPI, the Twilio SDK and the OpenAI APIs.

Alongside the employed work there is a shipped-product record that is unusual for an AI engineer: a multi-tenant SaaS CRM where every customer gets a physically separate database and its own running process, an SEO command centre driven by a set of specialised agents rather than one prompt, and a convolutional model trained end to end for chest X-ray classification. These are built systems with real users, not weekend prototypes.

• SuperManager (2026) — The agent and LLM layer of an enterprise agentic platform.
• NapCRM (2025) — Multi-tenant SaaS CRM where every customer gets an isolated database and its own process.
• Adnosaur (2024) — Product and ad-campaign discovery for dropshippers, with an AI-assisted store search.
• NapSEO (2025) — An SEO command centre run by a set of specialised agents rather than one prompt.
• Covix (2022) — A convolutional model that classifies chest X-rays for COVID-19 positivity.

Technical range

AI and ML

AI agents, LLM applications, RAG, NLP, Prompt engineering, OpenAI API, Computer vision, Deep learning

Languages

Python, JavaScript, TypeScript, Java, C++, PHP

Backend

FastAPI, Node.js, Express, Django, Flask, J2EE

Data

PostgreSQL, MySQL, MariaDB, MongoDB, Redis, Oracle

Platform

AWS, Docker, Kubernetes, CI/CD, Cloudflare, Git

Education

B.Tech, Computer Science and Engineering, I.K.G. Punjab Technical University, 2019 - 2023, Mohali, Punjab.

Availability

Open to senior AI engineering roles, and to consulting problems where the model is the easy part and the surrounding system is not. Email is the fastest route, and it is answered by a person rather than a form queue.