← Home
Partner Basecamp · Cohort 19

San Francisco

June 24–25, 2026
+55
Net Promoter Score
67
Day 1 Responses
98%
Matched Pairs
92%
Day 2 Response Rate
Programme Satisfaction
Strong promoter majority — NPS +55 reflects broad programme endorsement.
+55NPS · n=62
64% Promoters26% Passives10% Detractors
95% CI: +38 → +71  ·  True NPS lies within this range with 95% confidence (n=62 respondents)
Confidence Arc
Moderate end-of-Day-1 build confidence — cohort is progressing; Day 2 has room to close remaining gaps (mean 3.8/5).
3.8end-of-D1 build / 5
4.2D2 Design / 5
4.2D2 Commercial / 5
4.3D2 Build / 5
Audience
67 participants across 11 organisations — Developer majority with applying in practice the most common AI experience level.
Top Organisations
PwC (16) UST Global (13) McKinsey (9) Cognizant (9) Deloitte (6)
Voice of Participant
What participants said — dominant theme from each NPS segment.
Promoters
Hands-on, interactive format with practical AI techniques and real-world use cases resonated strongly with beginners and experienced professionals alike.
Passives
Advanced participants found optimization and context engineering valuable, but basics felt repetitive; governance and security gaps noted.
Detractors
Content delivery format inefficient for experienced audiences; some material better suited for asynchronous or remote consumption.
Survey & Data Quality
Survey & Data Quality
Day 1 responses67
Day 2 responses62 (92% of Day 1)
Matched pairs61 (98% of Day 2)
Orgs resolved67 of 67 respondents matched to named org
Day 2-only respondents1 (1.6%) — suppresses NPS by 2 points; no Day 1 data, excluded from persona & confidence analysis
Organisations
All Day 1 respondents · n=67
PwC16 (24%)
UST Global13 (19%)
McKinsey9 (13%)
Cognizant9 (13%)
Deloitte6 (9%)
Ascendion4 (6%)
AlixPartners3 (4%)
EPAM3 (4%)
Fractal Analytics2 (3%)
NEC1 (1%)
Lovelytics1 (1%)
Function × Seniority
All Day 1 respondents · n=67
FunctionPractitionerSr PractitionerMgr / Sr MgrDirector+Partner / MDTotal
Engineering81232126
Architecture2457–18
Business Leadership1–3217
Project / Engmt5–64116
Experience Profile
AI experience level · n=67 · % of row
By Persona
SegmentExploringApplyingDeliveringFrontiern
Developer19%50%8%23%26
Transformation Lead17%57%13%13%23
Architect17%33%28%22%18
By Seniority
SegmentExploringApplyingDeliveringFrontiern
Manager or Senior Manager24%35%29%12%17
Senior practitioner (5–9 years)12%44%25%19%16
Practitioner (0-4 years in role)25%62%—12%16
Director, Senior Director, or Principal13%53%7%27%15
Partner, Managing Director, or Executive—33%—67%3
How would you describe your current experience with AI tools prior to Basecamp?
Learning and exploring
12 (18%)
Applying in practice
32 (48%)
Delivering independently
10 (15%)
Operating at the frontier
13 (19%)
Prior to today, how much had you worked with Claude or the Anthropic API?
Not at all
5 (7%)
A little
26 (39%)
Regularly
36 (54%)
Did the depth land for this audience?
Technical depth perception · n=67 respondents · % of row
By Persona
SegmentToo basicAbout rightToo advancedn
Developer12%88%—26
Transformation Lead4%83%13%23
Architect11%89%—18
By AI Experience Level
SegmentToo basicAbout rightToo advancedn
Applying in practice6%91%3%32
Operating at the frontier23%77%—13
Learning and exploring8%75%17%12
Delivering independently—100%—10
Overall depth distribution
Too basic
6 (9%)
About right
58 (87%)
Too advanced
3 (4%)
Did the pace work across the room?
Session pace perception · n=67 respondents · % of row
By Persona
SegmentToo slowWell pacedToo fastn
Developer4%88%8%26
Transformation Lead9%78%13%23
Architect6%83%11%18
By AI Experience Level
SegmentToo slowWell pacedToo fastn
Applying in practice6%84%9%32
Operating at the frontier15%77%8%13
Learning and exploring—83%17%12
Delivering independently—90%10%10
Overall pace distribution
Too slow — could have covered more
4 (6%)
Well paced
56 (84%)
Too fast — not enough time to apply
7 (10%)
How confident were participants to build with Claude after Day 1?
End-of-Day-1 confidence · "How confident are you in your ability to build a client solution using Claude?" · n=67 · 1–5
Mean end-of-D1 build confidence by Persona
Developer3.92/5 · n=26
Architect3.72/5 · n=18
Transformation Lead3.65/5 · n=23
Programme mean: 3.9/5 · 15 cohorts
Mean end-of-D1 build confidence by AI Experience Level
Operating at the frontier4.54/5 · n=13
Delivering independently4.00/5 · n=10
Applying in practice3.72/5 · n=32
Learning and exploring2.92/5 · n=12
Overall end-of-D1 build confidence distribution
1
1 (1%)
2
8 (12%)
3
15 (22%)
4
24 (36%)
5
19 (28%)
How relevant was today's content to your current role?
Content relevance rating · n=67 respondents · 1 = not relevant, 5 = highly relevant
Mean relevance by Persona
Developer4.23/5 · n=26
Architect4.22/5 · n=18
Transformation Lead3.83/5 · n=23
Mean relevance by AI Experience Level
Delivering independently4.40/5 · n=10
Operating at the frontier4.38/5 · n=13
Applying in practice4.00/5 · n=32
Learning and exploring3.75/5 · n=12
Overall relevance distribution
1
0 (0%)
2
3 (4%)
3
10 (15%)
4
32 (48%)
5
22 (33%)
How likely are you to recommend attending this programme to a colleague?
n=62 Day 2 respondents · 92% of Day 1
Score distribution · 0 = not at all likely · 10 = extremely likely
Cohort012345678910
Cohort 19246102020
■ Detractors 0–6■ Passives 7–8■ Promoters 9–10
NPS Breakdown
64%
26%
10%
NPS +55 (n=62)
Promoters (9–10)Passives (7–8)Detractors (0–6)
Programme mean: +41 · 15 cohorts
Promoters mean: 51%
Passives mean: 40%
Detractors mean: 10%
NPS by Segment
Organisations, personas, and experience levels · ≥4 respondents shown
By Organisation
Promoters (9–10)Passives (7–8)Detractors (0–6)
Cognizant
89%
11%
0%
NPS +89 (n=9)
Deloitte
83%
17%
0%
NPS +83 (n=6)
Ascendion
75%
25%
0%
NPS +75 (n=4)
PwC
62%
31%
8%
NPS +54 (n=13)
McKinsey
62%
12%
25%
NPS +38 (n=8)
UST Global
46%
46%
8%
NPS +38 (n=13)
By Persona
Promoters (9–10)Passives (7–8)Detractors (0–6)
Architect
59%
41%
0%
NPS +59 (n=17)
Transformation Lead
68%
21%
10%
NPS +58 (n=19)
Developer
68%
20%
12%
NPS +56 (n=25)
Programme means · 15 cohorts: Architect +51 · Developer +46 · Transformation Lead +35
By AI Experience Level
Promoters (9–10)Passives (7–8)Detractors (0–6)
Applying in practice
70%
27%
3%
NPS +67 (n=30)
Delivering independently
60%
40%
0%
NPS +60 (n=10)
Learning and exploring
67%
11%
22%
NPS +44 (n=9)
Operating at the frontier
58%
25%
17%
NPS +42 (n=12)
Programme means · 15 cohorts: Learning and exploring +44 · Applying in practice +43 · Delivering independently +40 · Operating at the frontier +37
What is the main reason for your score?
n=49 responses · organised by NPS segment
Promoters (score 9–10)· 34 responses
Hands-on, interactive format with practical AI techniques and real-world use cases resonated strongly with beginners and experienced professionals alike.
10“This was very interactive and to the point session. All our apprehensions on how to get AI work effectively for our client were resolved. We are definitely leaving the session with lot of information that will help boost our business as well as collaboration with Anthropic”
Passives (score 7–8)· 10 responses
Advanced participants found optimization and context engineering valuable, but basics felt repetitive; governance and security gaps noted.
8“Training is very good, especially the modules on the second day around inference optimization and context engineering. However, I think the rest is relatively basic if you are someone with AI experience. Useful reminders, but giving an 8 for that reason.”
Detractors (score 0–6)· 5 responses
Content delivery format inefficient for experienced audiences; some material better suited for asynchronous or remote consumption.
6“Many things could have been done remotely and offline.”
End-of-Programme Confidence
n=62 respondents · 1 = not confident, 5 = very confident
How confident are you in your ability to design an evaluation for an AI solution?
12345Mean
——21%44%35%4.15
Programme mean: 4.1/5 · 15 cohorts
Leaving today, how comfortable are you advising clients on AI when questions arise?
12345Mean
—2%13%53%32%4.16
Programme mean: 4.2/5 · 15 cohorts
How confident are you in your ability to lead a conversation about Anthropic and Claude with a client?
12345Mean
——13%47%40%4.27
Programme mean: 4.2/5 · 15 cohorts
Confidence Trajectory by Persona
D1→D2 delta · apply-AI confidence · matched pairs only · n=61
Positive delta = confidence grew · negative = dropped
Architect (n=17)
+0.65
Developer (n=25)
+0.36
Transformation Lead (n=19)
+0.42
What one takeaway will you share with a colleague or client?
n=47 responses
Participants value learning optimal Claude usage patterns and model selection rather than basic tool familiarity.
“Don't just use Claude. Use Claude the optimal way and watch the magic”
Participants want systematic evaluation frameworks to validate AI outputs rather than relying on intuition or assumptions.
“Iterate on Evals not on intution”
Participants seek practical strategies for different AI implementation stages and compliance considerations for client delivery.
“Power of AI, different stages and strategies”
Participants value rapid, practical agent-building skills that directly improve team productivity and task completion speed.
“Learned how to build agents in under an hour, improving developer productivity and helping teams complete tasks faster.”
Participants prioritize interactive exercises and practical experimentation over theoretical instruction or expensive model usage.
“Exercises and not everything has to be ran with opus :)”
What will you build for yourself or apply at work in the next 30 days?
n=48 responses
Participants want to build specialized agents for knowledge management and business productivity tasks like presentation generation.
“Knowledge management agents”
Participants are actively building multi-agent systems with evaluation capabilities for their engineering and CAD tool clients.
“already building a multi agent multi skill cad tool for engineering companies”
Participants seek production-ready patterns for securing Claude in development workflows and spec-driven architecture approaches.
“how to securely use claude for development and ready for the production deployment”
Participants identify security and cost modeling as critical gaps when applying AI to industry solutions and integrations.
“Apply what is learnt. Still there are gaps on security and costing part of it.”
Participants recognize evaluation tools and skill testing as essential infrastructure for validating agent performance.
“Conversion tools with evals”
Partner Basecamp Survey Report · Cohort 19 · Generated by analyze.py v2.3
NPS Reason
n=49 responses · grouped by NPS segment
Promotersscore 9–10 · 34 responses
1Good lessons10
2Good intro. To Agentix programming10
3Enjoyable and helped me learn a lot of applicable techniques - and great food10
4Learnt a lot10
5Very informative10
6Relevant for the AI beginner10
7As a product manager and engineer this was incredibly helpful for my job and career. Affirmed my current ways of working and new ways/approaches to embed into my career and day to day.10
8Really like the hackathon10
9New ideas , new learnings, getting evolved with AI10
10Excellent immersion concept10
11It’s very helpful and loved the way we worked together10
12This was very interactive and to the point session. All our apprehensions on how to get AI work effectively for our client were resolved. We are definitely leaving the session with lot of information that will help boost our business as well as collaboration with Anthropic10
13The content was very informative and the atmosphere is fantastic10
14Got very good understanding of capabilities of Anthropic Claude Models and the best practices to follow while developing agents etc10
15Covers both business and technology with ease and in depth10
16Very hands on10
17valuable insights and learnings on not just how to run but optimize and maintain ai processes10
18Hands on practice with pre defined examples9
19Very very informative and engaging9
20Interactive and lots of hands on9
21Content related to ai use cases9
22Content9
23Good interactive and facilitated for ai work environment9
24Not a ten just because it’s a lot for two days9
25Great learnings, great devlolment tools, great speakers9
26Very practical and thought provoking use cases and practical training9
27The hackathon was really helpful. I would definitely like to see more collaborative activities like the hackathon, as the other activities were way more individual.9
28Would recommend.9
29It’s a great overall experience and learning to understand and start the journey.9
30Good explainer on anthropic development philosophy and how to sell9
31Informative9
32Super hands on interactive learning with strong presenter and host support Great tunes the entire time Too9
33Good combination of learning, hands on creation, hackathon, and Q&A. Training went over a lot of important concepts when it comes to incorporating AI in the daily workflow to create something complex.9
34I learnt a lot. The exercises were great. Environment was appropriate for learning. Interactive sessions. And opportunity to mix. Food was great.9
Passivesscore 7–8 · 10 responses
1Super fun and technical, learned a lot8
2Good starter to understand Claude capabilities8
3Grounded one for new entry to Claude.8
4Good overview of Claude capabilities. Hands on experience was good.8
5Training is very good, especially the modules on the second day around inference optimization and context engineering. However, I think the rest is relatively basic if you are someone with AI experience. Useful reminders, but giving an 8 for that reason.8
6New features7
7it is leaning towards a lively environment to refresh and learn together7
8governance and security can also be part of the module7
9More details on optimization needed.7
10I like the degree of hands on, would have been good to do a demo walkthrough of each exercise before starting each session.7
Detractorsscore 0–6 · 5 responses
1Helpful but confusing at parts6
2New tech awareness6
3Many things could have been done remotely and offline.6
4Knowing Claude better and applying6
5Most of my colleagues know this information already5
Most Valuable
What one takeaway will you share with a colleague or client? · n=47 responses
1Agentic approaches
2Na
3Val’s frame works
4Claude routine
5Don’t just use Claude. Use Claude the optimal way and watch the magic
6Eval frameworks and strategies.
7Different models for different purposes
8Dont miss
9Human in the loop will never be ended
10Claude
11Iterate on Evals not on intution
12i will create the document for reporting.
13Inference optimization options
14Great learning opportunity and hands on experience
15Power of AI, different stages and strategies
16Start using loops and when you’re stuck and have a question it’s simple, ask Claude
17Be smart about using AI and always evaluate the work
18Claude features are good and useful in day to day developer work
19Evaluations are critical for the success of many implementations.
20This is actually complicated.
21The power of skills in anthropic
22in short how to use AI properly
23evals are most important
24It’s magic and it’s different
25About claude compliance AI and how we can provide more use case to improve it
26The value of Agentic Evaluations
27With Claude you can accelerate the path to value.
28You don’t always need the fanciest model to do what you want to do with AI.
29Gained good knowledge on Claude will share it with clients
30Eval framework playbook
31Exercises and not everything has to be ran with opus :)
32The effectivenssof evals
33Cost optimization is worth the effort
34How world is moving to no mans code
35Claude code in desktop is the best application of Claude for non technical roles
36Demo
37Good mix of tech and product
38Evals process and how to refine them
39When a model hallucinates, it’s not the model but the prompt
40Cost optimization strategies
41Using AI needs mindset shift and there are a variety of tools at disposal especially via Anthropic that really can elevate someone
42Context usually matters more than the model. Respect the principal of weakest model to limit token usage.
43Learned how to build agents in under an hour, improving developer productivity and helping teams complete tasks faster.
44It felt very accessible for all levels and gave us resources and tips we could immediately implement
45Anthropic really believes in providing value to enterprises.
46harnesses
47Great program for developers, somewhat mis-cast as a program for business development. Content was great, could use some leveling up.
30-Day Intentions
What will you build for yourself or apply at work? · n=48 responses
1More Ai products
2Specialist Agents
3Knowledge management agents
4Agent to built great power point slides
5Coding skill
6Im going to build many new skills that package reusable skills
7Jira Agent
8ProjeX - Agentic PMO
9I am planning to build an application for my kids activity finder
10Conversion tools with evals
11Using routines and advisors
12A karaoke agent
13AIOps automation
14Result oriented high performing Agentic solutions using Claude.
15Apply AI for Integrations and Industry solutions
16Yes definitely
17iOS app for personal use and enterprise ai solutions
18Probably a transformational product in data space
19Apply what is learnt. Still there are gaps on security and costing part of it.
20Will start on it
21Not sure
22Learning and development tools
23Agentic AI
24already building a multi agent multi skill cad tool for engineering companies
25how to securely use claude for development and ready for the production deployment
26Continue the excellence with claud
27Cybersecurity VM agent
28I’m planning to build a POC for an app on Canada’s newly launched Open Banking Platform
29I have already built a harness around Claude for a spec driven development using multiple archetypes
30I am going to try rebuilding agents I built on ChatGpt on Claude. I like the usability and integration into my workflow so much better.
31Build a deliverable full end to end solution
32Building training materials for a client to drive adoption.
33I want to build a video game myself but I want to build swarm agents to be able to cost optimize
34Application on the cyber security space as well as aisldc
35MCP tools for agentic loops to resolve manual internal processes
36Im my project to integrate 3rd party app
37Executive reporting dashboard to monitor the PWX x ANT alliance
38More apps
39Improve product development workflows.
40Build an agents using API to build faster processes
41Probably
42Something this training helped remind me of is all the new Claude Code offering like dynamic workflows and agent teams. I will be exploring those in the next month.
43Everything learned during the base camp is extremely applicable in real life scenario - both personal and client
44Project POCs and personal projects
45I want to build a client hunting swarm
46Will build some optimization tool
47tests for skills
48Several opportunities with client engagements I lead.