Skip to content

Commit 8694dc0

Browse files
claudeanandgupta42
authored andcommitted
Enrich design doc with OpenClaw research and proactive behaviors
Add detailed competitive analysis from OpenClaw (self-improving memory, heartbeat scheduler, meet-users-where-they-are), Devin ($10.2B valuation, "junior partner" framing), and Factory AI (workflow embedding). Add proactive behaviors section with background monitors (cost alerts, freshness checks, schema drift, PII scanning) and auto-promotion of learned corrections. https://claude.ai/code/session_01V17Kk3qCZFp9ZJiuNYucoq
1 parent 01225b7 commit 8694dc0

1 file changed

Lines changed: 108 additions & 12 deletions

File tree

docs/design/ai-teammate-repositioning.md

Lines changed: 108 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -17,7 +17,26 @@ The "AI teammate" framing creates a different mental model entirely:
1717

1818
### Inspiration: OpenClaw & the "Trainable Agent" Pattern
1919

20-
OpenClaw and similar projects (SWE-agent, Devon, etc.) demonstrate that the most powerful AI agents aren't the ones with the most tools — they're the ones that **learn from their environment**. The key insight:
20+
**OpenClaw** (247K+ GitHub stars, fastest-growing open-source project ever) proved the "teammate" framing works when backed by real architecture. Key lessons:
21+
22+
1. **Meet users where they are.** OpenClaw's UX *is* your existing messaging apps (WhatsApp, Telegram, Slack, Signal). Zero learning curve. For altimate, the equivalent: meet data engineers in their terminal, their dbt workflow, their Slack — don't force them into a separate app.
23+
24+
2. **Self-improving memory.** OpenClaw captures learnings, errors, and corrections in structured files (`LEARNINGS.md`, `ERRORS.md`). When patterns recur 3+ times across 2+ tasks within 30 days, they auto-promote into permanent system prompt files (`CLAUDE.md`, `SOUL.md`). This is the model for altimate's training system — learning should be automatic, not manual.
25+
26+
3. **Proactive heartbeat.** A scheduler wakes the agent at intervals so it can act without being prompted — checking email, running tasks, flagging issues. For altimate: imagine the teammate running nightly cost checks, freshness monitors, or schema drift detection without being asked.
27+
28+
4. **Persistent identity.** One agent instance across all channels with shared memory and context. For altimate: the same teammate across TUI, web, CI/CD, and Slack — always knowing your project, your standards, your history.
29+
30+
**Devin** ($10.2B valuation, $73M ARR) proved the market appetite: they market as "a collaborative AI teammate" and "the first AI software engineer," but candidly advise treating it as "a junior coding partner." The honesty works — users understand the capability boundary.
31+
32+
**Factory AI** positions autonomous "Droids" that embed into existing workflows (VS Code, JetBrains, Slack, Linear). Their insight: "delegate complete tasks like refactors, incident response, and migrations without changing your tools."
33+
34+
The **World Economic Forum** outlines the industry evolution:
35+
1. **Copilots** (assisted intelligence) — suggestions, human controls
36+
2. **Agents** (autonomous task execution) — limited decisions, task-oriented
37+
3. **AI Teammates** (collaborative intelligence) — adapt, learn, achieve shared objectives
38+
39+
altimate should skip straight to level 3 for data engineering.
2140

2241
> The best AI teammate is one that adapts to YOUR team, not one that forces your team to adapt to IT.
2342
@@ -100,7 +119,7 @@ patterns:
100119
- naming: snake_case
101120
```
102121
103-
#### b) Learn-by-Correction (`/feedback`)
122+
#### b) Learn-by-Correction (Implicit + `/feedback`)
104123

105124
When the teammate does something wrong, instead of just fixing it, you teach it:
106125

@@ -125,6 +144,18 @@ applies_to:
125144
severity: high
126145
```
127146

147+
**Auto-promotion (inspired by OpenClaw's self-improving agent):**
148+
149+
When a correction pattern recurs 3+ times across 2+ sessions within 30 days, it auto-promotes from a "soft preference" to a "hard rule" in the system prompt. The teammate proactively tells you:
150+
151+
```
152+
Teammate: I've noticed you've corrected my use of LEFT JOIN → INNER JOIN
153+
in staging models 4 times now. Should I make this a permanent rule?
154+
[Yes, always use INNER JOIN in staging] [No, keep case-by-case]
155+
```
156+
157+
This mirrors OpenClaw's `LEARNINGS.md` → `CLAUDE.md` promotion pipeline, but adapted for structured data engineering rules rather than free-form text.
158+
128159
#### c) Learn-by-Document (`/train`)
129160

130161
For more formal training, point the teammate at documents:
@@ -315,7 +346,54 @@ Teammate: Learned! This pattern is saved to .altimate/learned/patterns/staging-m
315346
Commit this file so your teammates' altimate instances learn it too.
316347
```
317348
318-
### 6. The Training Dashboard (`/training-status`)
349+
### 6. Proactive Behaviors (inspired by OpenClaw's Heartbeat)
350+
351+
A real teammate doesn't wait to be asked. OpenClaw proved this with its heartbeat scheduler — a background process that wakes the agent at intervals to act autonomously.
352+
353+
For altimate, proactive behaviors are domain-specific and high-value:
354+
355+
#### Background Monitors (opt-in)
356+
357+
```yaml
358+
# .altimate/monitors.yml
359+
cost_alert:
360+
schedule: daily
361+
threshold: 120% # Alert if costs >120% of 30-day average
362+
action: slack_notify # or: session_note, email, github_issue
363+
364+
freshness_check:
365+
schedule: hourly
366+
models: ["fct_orders", "fct_revenue", "dim_customers"]
367+
stale_after: 4h
368+
action: slack_notify
369+
370+
schema_drift:
371+
schedule: daily
372+
watch: ["raw.*"] # Watch raw layer for upstream changes
373+
action: session_note # Note it in next session
374+
375+
pii_scan:
376+
schedule: weekly
377+
scope: all_indexed_schemas
378+
action: github_issue
379+
```
380+
381+
#### In-Session Proactive Behaviors
382+
383+
```
384+
You: Can you build a staging model for the new payments table?
385+
386+
Teammate: Sure! Before I start — I noticed something while looking at the
387+
payments table:
388+
- Column `card_number` matches PII pattern (credit card)
389+
- Column `email` matches PII pattern (email address)
390+
These should probably be masked or excluded in staging.
391+
Want me to add masking, or exclude them?
392+
```
393+
394+
The teammate doesn't wait to be asked about PII — it checks proactively because that's what a good data engineering teammate does.
395+
396+
### 7. The Training Dashboard (`/training-status`)
319397

320398
Show users what their teammate has learned:
321399

@@ -444,17 +522,35 @@ Want to review or modify any learned patterns? Use /teach --list
444522

445523
## Competitive Differentiation
446524

447-
| Product | Framing | Training? | Data-Aware? |
448-
|---|---|---|---|
449-
| Claude Code | AI coding assistant | No (just prompts) | No |
450-
| Cursor | AI-powered IDE | Cursor Rules files | No |
451-
| Devin | AI software engineer | No | No |
452-
| OpenClaw | Trainable AI agent | RL from feedback | No |
453-
| **altimate** | **AI data teammate** | **Yes (/teach, /train)** | **Yes (55+ tools, warehouse)** |
525+
| Product | Framing | Training? | Data-Aware? | Proactive? |
526+
|---|---|---|---|---|
527+
| Claude Code | AI coding assistant | CLAUDE.md only | No | No |
528+
| Cursor | AI-powered IDE | Cursor Rules files | No | No |
529+
| Devin ($10.2B) | AI software engineer | No | No | Yes (async tasks) |
530+
| Factory AI | Autonomous Droids | No | No | Yes (workflow triggers) |
531+
| OpenClaw (247K stars) | Trainable AI agent | Self-improving memory + RL | No | Yes (heartbeat scheduler) |
532+
| **altimate** | **AI data teammate** | **Structured learning (/teach, /train, auto-promote)** | **Yes (55+ tools, warehouse)** | **Yes (cost alerts, schema drift)** |
454533

455-
The unique combination: **trainable + data-domain-specific + warehouse-connected**.
534+
### What altimate takes from each:
456535

457-
No other product lets you teach an AI your team's SQL standards and then have it enforce those standards with direct access to your warehouse metadata, lineage, and cost data.
536+
| From | What we borrow | How we adapt it |
537+
|---|---|---|
538+
| **OpenClaw** | Self-improving memory, auto-promotion of learnings | Structured YAML rules instead of free-form markdown; domain-specific (SQL patterns, not general tasks) |
539+
| **OpenClaw** | Heartbeat scheduler for proactive behavior | Nightly cost checks, freshness monitors, schema drift detection |
540+
| **OpenClaw** | Meet-users-where-they-are UX | TUI + Web + Slack + CI/CD — same teammate everywhere |
541+
| **Devin** | "Collaborative AI teammate" positioning | Same framing, but specialized: "data engineering teammate" not "software engineer" |
542+
| **Devin** | Honest capability framing ("junior partner") | "Trained on your standards, but you're still the senior engineer" |
543+
| **Factory AI** | Embed into existing workflows, don't replace them | Works inside your dbt workflow, not beside it |
544+
545+
### The unique combination
546+
547+
**Trainable + data-domain-specific + warehouse-connected + proactive.**
548+
549+
No other product lets you:
550+
1. Teach an AI your team's SQL standards (`/teach`)
551+
2. Have it enforce those standards against your actual warehouse metadata and lineage
552+
3. Watch it auto-improve from your corrections over time
553+
4. Wake up to find it already flagged a cost anomaly or schema drift overnight
458554

459555
---
460556

0 commit comments

Comments
 (0)