Developing agentic software?
I am posting content here that will help you get idea on building agentic software systems.
Before you proceed, you need to remember…
Machines don’t learn or think like humans do! It’s all math.
In my opinion, crafting agentic software depends mainly on five maturity areas…
1️⃣ Complexity of the problem and how agents can help (will it really benefit with such a solution?)
2️⃣ The level of accuracy and human intervention required (beware, LLMs are probabilistic software!)
3️⃣ Agent personas with clear objectives: Write clever prompts using different prompt engineering techniques
4️⃣ Thinking simple workflows: Restrospecting the output at every step, understand what can improve accuracy
5️⃣ Implications on security, privacy and compliance
To build capabilities in agentic AI, you could use this three step process.
I followed this three step process myself.
✨ Step 1: Be an efficient software craftsman
✅ Stop thinking that AI is the answer to all your problems
✅ Be part of, or develop an excellent engineering culture
✅ Collaborate, experiment and learn; understand what others are doing
✅ Learn Python, async programming, API development and one database
✅ Read why LLMs are called probabilistic systems, use ChatGPT or Gemini and experiment with prompts
✨ Step 2: Understand the building blocks
✅ Start small; build a simple chatbot and experiment with different prompt techniques
✅ Learn vector databases, tokens, text embedding, chunks, etc
✅ Build a RAG workflow, connect it with the chatbot, add memory - Try Agentic RAG!
✅ Learn different RAG techniques and simple tool usage with LLMs
✅ Understand the economics of cloud computing
✨ Step 3: Develop agents
✅ Build a system with two agents that interact with each other, define agent goals
✅ Optimize tokens by engineering agent context!
✅ Learn structured inputs, structured outputs and guardrails
✅ Learn agentic workflow patterns and orchestration, add multiple agents
✅ Provide tools to the agents, learn Model Context Protocol
✅ Logging, monitoring, LLMOps and deployment
✅ Dive deep into evals (Evaluation of how the agents are performing!) - LLM as judge is a popular technique.