You'll want to preserve the option to switch between models on your AI apps. Eval frameworks enable switching confidence.

Don't rush to fully autonomous AI apps. Use a human in the loop to get product out faster and learn with more depth.

Inference costs of the LLM API is an important factor in building AI apps; especially for young startups. Open source models (like ones hosted by groq), eval frameworks, and observability make this a lot easier.

Don't AI just to AI. Traditional software still has a place, saving your AI resources for the real use cases.

Integrating AI apps into traditional systems unlocks a lot of value, but becomes much harder. Writing quality applications that integrate well is a key. Austin Vance shares strategies to make that seamless.

Chat isn't the best use of GenAI and LLMs but it's the commonly used medium.

The safety and confidence from eval frameworks are critical for moving fast, switching architecture, and adding new features on your Ai apps.

DSPy simplifies prompt tuning for optimal LLM responses. We fine-tune prompts based on input/output analysis, addressing incorrect or inappropriate LLM behavior.

LLM Observability enables Build, Deploy, Learn for AI Apps

Keyword search vs Thematic search. Austin Vance talks to us about how his team is building agentic apps on LLMs.

Repetitive knowledge work is a ripe candidate for a disurptive agentic app

Eval frameworks give similar value of a test suite - giving you confidence to change your application with impunity.

While assuming positive intent is a virtue, we unfortunately encounter coworkers who aren't honest on occasion. Learn more about managing inauthentic coworkers, and navigating uncomfortable conversations from our episode with Joshua Graves.

Perform better during stressful situations by understanding the amygdala and rewiring your behaviors can help. Learn more about overcoming stress, and navigating uncomfortable conversations from our episode with Joshua Graves.

Silence is expensive and often compounds: Netflix's Communicatioh Secret Learn more about overcoming stress, and navigating uncomfortable conversations from our episode with Joshua Graves.

Unlock your team's instincts and creativty by modelling vulnerability, and fostering psychological safety Learn more about psychological safety, and navigating uncomfortable conversations from our episode with Joshua Graves.

User manuals for humans? Foster ease of communication and collaboration by sharing your user manuals with your team. Learn more about counterintuitive collaboration, and navigating uncomfortable conversations from our episode with Joshua Graves.

Team rituals are massive unlocks, especially when they are intentional, pragmatic, and co-created by the entire team (not just the leader). Learn more about overcoming stress, and navigating uncomfortable conversations from our episode with Joshua Graves.

Just like tech debt, diluting your feedback loops when things get busy come with a cost. Learn more about overcoming stress, and navigating uncomfortable conversations from our episode with Joshua Graves.

Rituals at work, despite being 'religious-informed' have benefits like meaning, familiarity, and opportunities for growth. Similar to religion, it maybe worthwhile to be careful if you find yourself being dogmatic or meaningless.