TR
04 · 2 hours

Research First, Then Decide

Multi-agent research, a reviewer agent, tokens and cost.

Request this programmeDownload the catalogue (PDF, Turkish)

Goal: Participants prepare an important decision not in a single chat but with a workflow of agents that research, challenge and review, while keeping the cost under control.

Who it is for

  • Managers who prepare decisions
  • Strategy, business development and project management teams
  • Advanced users who rely heavily on AI

What participants gain

  • A research and review workflow built for their own work
  • A decision note template
  • An installed Context Mode and a cost checklist

Content

  1. 01The limit of a single chatHow context gets lost in long, multi-part work, and the risk of a single point of view.
  2. 02Working with several agentsAgents that research the same question from different angles in parallel, and how to design the workflow.
  3. 03The reviewer agentAn independent second opinion that tries to disprove the result.
  4. 04The decision noteBringing findings, uncertainty and the recommendation together on one page.
  5. 05Tokens and costHow cost builds up, why the context fills, and why Turkish text uses more tokens.
  6. 06Saving with Context ModeProcessing large outputs without loading them into the context, with an open source tool on GitHub; installation.
WHAT YOU KEEP
A research and review workflow and an installed Context Mode
Duration
2 hours. Can be taken on its own; two programmes back to back make a half day.
Participants
Up to 10 people, each with their own laptop.
Data
No company data is used. Exercises use open data and non-sensitive content.
Prerequisites
No coding is required. The installation list is shared in advance and approved by IT.

Other programmes

All programmes

LET’S START

For most organisations, training is the first step.

Once your team has learned, if you want to turn ideas into a pilot, we continue with AI transformation and roadmapping.

Request a meetingSee AI transformation