Every year, we review our tool stack. 2026 was a year of consolidation: fewer tools, but better integrated. Here are the 8 tools we use daily at DataSAI, with our honest opinion on each.
Our tooling philosophy
We stopped chasing new tools every week. A tool enters our stack only if it replaces 2 existing tools OR unlocks a previously impossible use case.
The 8 essential tools
1. Cursor: the AI code editor
Cursor replaced VS Code in our team. The Claude and GPT-4 integration accelerates development by 30-50%, especially for navigating existing codebases.
2. LangSmith: agent monitoring
Impossible to do serious production work without LangSmith. Every LLM call traced, cost per request, latency, error rate.
3. Supabase: the AI database
PostgreSQL with pgvector extension, built-in auth, auto-generated API. Our default choice for any AI project requiring storage.
4. Langfuse: the open source alternative
For projects where data cannot leave the client's infrastructure: open source, self-hostable, functionally equivalent to LangSmith.
5-8: dbt, Metabase, n8n, Unsloth
dbt for versioned SQL data transformation. Metabase for non-technical dashboards. n8n for no-code automation. Unsloth for open source LLM fine-tuning.
Our advice: master 5 tools deeply rather than knowing 20 superficially. Value is in the combination and integration.
With care,
Excellent article, this matches exactly what we're seeing with our enterprise clients. The section on inference costs is especially valuable. It's a topic most articles gloss over but it's make-or-break at scale.
Thanks James! Inference cost optimization is often deprioritized during prototyping but becomes critical in production. Feel free to book a session if you'd like to go deeper on this.
Sharing this with my whole team. The distinction between an impressive demo and robust production is exactly the debate we're having internally right now. The human checkpoint advice is immediately actionable.
Great article. I'd push back slightly on the 18-day deployment estimate, in our experience with enterprise security and GDPR requirements, 4–6 weeks is more realistic for a first production agent.
Completely fair point David. The 18 days refers to a scoped first agent in a test environment. For full enterprise production with security constraints, your estimate is accurate.