At the 5 September Hyderabad meetup, “In the Loop,” we explored agent harnesses, enterprise agents, and what makes AI output good enough for its intended use. Attendees responded positively to the talks and the overall experience, found the technical level well pitched, and showed enthusiasm for contributing to future meetups.

Alongside that encouragement came a clear request: more demos, working examples, and practical implementation detail. Agents in production and applied AI topped the list of topics attendees wanted next, followed by fine-tuning, RL environments, and small language models.

That feedback shapes our next meetup, “After the Prototype,” on 28 November. We want to hear what you built, what happened when people used it, and what you learned along the way.

We’re inviting proposals across three areas:

Agents in production: assistants, coding agents, voice bots, and workflow automation. What broke under real traffic? What did you build around the agent to make it reliable? What did it cost in latency, money, and engineering time?
Applied AI: RAG, MCP, skills, plugins, and agent tooling built into products or shared team workflows. What didn’t work in the first version? What did you have to change or build yourself?
Fine-tuning, RL environments, and small language models: what made you choose this approach, how you evaluated it, and what you gained or gave up in quality, speed, and cost.

The decisions, failures, and fixes behind your system are often the parts others can learn the most from—and the parts that rarely make it into documentation or demos. Sharing them helps fellow engineers avoid dead ends and make better choices. It also gives you a chance to put your approach in front of peers, get feedback on difficult trade-offs, and connect with people tackling similar problems.

You don’t need a finished slide deck to submit. Start with an abstract or mind map outlining the problem, your approach, and what attendees can learn. Work in progress and experiments you decided not to ship are welcome too. A work in progress with real findings, or an experiment you decided not to ship, can make a useful talk. Bring enough detail for others to learn from it: an architecture walkthrough, code example, demo, or a clear account of what you tried and why you changed course.

Our review panel will be glad to work with you through the review process to sharpen your pitch and shape your talk.

Submit your proposal by 31 October: https://hasgeek.com/fifthelephant/hyd-meetup-cfp/

After the Prototype - agents, applied AI, and models put to work

{{ gettext('Draft') }}