AI Agents Are the New Software Engineers — Here's How They Actually Work
#computer-science
#file-systems
#learning
#ai
AI agents have evolved far beyond chatbots and code autocomplete. Today's agentic systems can plan, write, test, and debug software autonomously — functioning as active participants in the development lifecycle. But what's actually happening under the hood?
This session breaks down the architecture behind AI coding agents: how large language models are augmented with tool use (file systems, terminals, code search, compilers) to interact with real codebases, how they decompose complex tasks into executable steps, and how the agentic loop — plan, act, observe, adapt — drives multi-step reasoning. We'll examine the role of context windows, retrieval-augmented generation, and guardrails that keep agents reliable and safe.
We'll then explore the system design and architectural patterns that make agents work at scale — orchestration frameworks, memory management strategies, multi-agent coordination, and the tradeoffs between autonomy and control. Attendees will gain insight into how these systems are designed end-to-end, from prompt engineering and tool integration to feedback loops and failure recovery.
Critically, this session reframes the narrative: AI agents are not here to replace engineers — they're force multipliers. The graduates who thrive won't be the ones competing with agents, but the ones who know how to direct them. We'll discuss practical ways students can prepare now — building strong fundamentals in system design and problem decomposition, learning to evaluate and guide AI-generated output, and developing the judgment that agents still lack. The engineers of tomorrow won't be measured by how fast they type code, but by how effectively they architect solutions with intelligent tools at their side.
Whether you're studying computer science, cybersecurity, or IT — understanding how these systems work isn't optional anymore. It's the new literacy.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
Loading virtual attendance info...
- Starts 03 April 2026 04:00 AM UTC
- Ends 22 May 2026 04:00 AM UTC
- No Admission Charge
Speakers
Harshit Kohli
Topic:
AI Agents Are the New Software Engineers — Here's How They Actually Work
Biography:
Harshit Kohli is a Senior Technical Account Manager at Amazon Web Services (AWS) with over 15 years of experience in IT infrastructure, cloud architecture, and enterprise solution design. He specializes in cloud modernization, advanced analytics, and deploying artificial intelligence and generative AI workloads at scale. At AWS, he partners with enterprise customers to deliver secure, resilient architectures, lead complex technical engagements, and mentor teams to improve operational excellence. Harshit is pursuing a Doctorate in Artificial Intelligence at the University of the Cumberlands and holds an MS in Information Technology and a BTech in Electronics and Instrumentation. He has authored multiple IEEE AI research papers, served as a conference chair and reviewer, judged industry hackathons, and is a published author in AI-driven healthcare systems.
Email:
Address:United States