Best on desktop, but the demo still works on mobile.

← Scorecard

Item 21 · synthetic

Synthetic Ambig Ai In Dev Only

scenario synthetic_ambig_ai_in_dev_only

Input

Tobi Adekunle
VP Product at Stackveil

I've been in developer tooling long enough to know that the hardest part isn't building the product — it's making sure engineers actually trust what it's telling them. That's the problem I came to Stackveil to work on.

We build distributed tracing and log aggregation infrastructure for backend engineering teams at mid-sized SaaS companies. My day-to-day is split between roadmap, working closely with our three PMs, and spending a lot of time with customers who are drowning in alert noise and false positives.

We closed our Series A about eighteen months ago, which gave us the runway to start doing some genuinely interesting things on the product side. Right now I'm piloting an anomaly detection layer that uses ML to surface meaningful signal from trace data — the goal is to cut time-to-detection for P1 incidents without requiring teams to hand-tune thresholds manually. We're in a private beta with six design partners, and the feedback has been humbling in the best way.

Before this, I was a PM at Datadog for four years, then spent two years leading product for a smaller APM startup that got acqui-hired. That background shapes how I think about instrumentation UX and the gap between what observability tools promise and what on-call engineers actually experience at 2am.

I'm based in London. Happy to connect with anyone working on the intersection of developer experience and applied ML.

Stackveil is a London-based B2B SaaS company providing distributed tracing, log aggregation, and alerting infrastructure for backend engineering teams. Founded in 2021 and backed by a Series A round closed in early 2023, the company serves mid-market SaaS businesses seeking to reduce observability complexity and alert fatigue. Stackveil employs approximately 55 people across engineering, product, and go-to-market functions. Its platform integrates with common cloud-native stacks and positions against larger players on configurability and lower instrumentation overhead. An AI-driven anomaly detection feature is currently in private beta with design partners.

Gold

Expected action
propose
Fit score
0.78
Industry
B2B SaaS
Segment
Observability / distributed tracing
Seniority
VP
Company size
51-200
Language
en
Notes
AI is in active development (private beta), which satisfies the ICP product_shape predicate at 1.0. Soft spot is arr_match — Series A observability at 55 ppl likely has ARR near or just under the $2M floor.

Predictions

Integratedmiss

  • action auto_add, expected propose
Action
auto_add ✗ (gold: propose)
Fit
0.85 (gold 0.78, |Δ| 0.07)
Industry
B2B SaaS ✓
Segment
Developer tooling / Observability ✗
Seniority
VP ✓
Company size
51-200 ✓
Grounding
8/8 claims grounded (100.0%)
Claims (8)
  1. Tobi holds the VP Product title at Stackveil, an explicitly listed ICP target role.
    ✓ in input VP Product at Stackveil
  2. Stackveil is a B2B SaaS company, matching the required product shape.
    ✓ in input Stackveil is a London-based B2B SaaS company
  3. Stackveil closed a Series A in early 2023, placing it within the ICP's target funding stages.
    ✓ in input Series A round closed in early 2023
  4. Stackveil has approximately 55 employees, well within the 20–250 ICP headcount range.
    ✓ in input Stackveil employs approximately 55 people across engineering, product, and go-to-market functions
  5. Stackveil has an AI feature actively in development — an ML-driven anomaly detection layer currently in private beta.
    ✓ in input An AI-driven anomaly detection feature is currently in private beta with design partners
  6. The ML feature is designed to reduce manual threshold tuning and cut time-to-detection for P1 incidents.
    ✓ in input an anomaly detection layer that uses ML to surface meaningful signal from trace data — the goal is to cut time-to-detection for P1 incidents without requiring teams to hand-tune thresholds manually
  7. Tobi's core customer problem is alert noise and false positives in observability tooling.
    ✓ in input spending a lot of time with customers who are drowning in alert noise and false positives
  8. Tobi brings deep domain context from four years as a PM at Datadog before joining Stackveil.
    ✓ in input I was a PM at Datadog for four years
Draft hook

Hi Tobi — the work you're doing at Stackveil resonates immediately. You're piloting an anomaly detection layer that uses ML to surface meaningful signal from trace data for customers who are drowning in alert noise and false positives — that's a genuinely hard product problem, and you're solving it in exactly the way we see the best teams approach it. I'd love to share a perspective on how others in private beta with similar AI features have accelerated the path from design partner feedback to GA. Worth a 20-minute call?

Reasoning

All five ICP dimensions are assessable from the input. Role match is perfect — Tobi is explicitly VP Product, a named ICP target role. Headcount match is strong at ~55 employees (ICP: 20–250). Stage match is strong at Series A (ICP: A/B/C). Product shape match is the highest-conviction signal: Stackveil is a B2B SaaS company with an ML-driven anomaly detection feature actively in private beta with design partners, satisfying the 'shipping or in active development' AI feature criterion. The only dimension with meaningful uncertainty is ARR — no figure is mentioned anywhere in the input — so arr_match is scored at 0.50 to reflect that it is plausible but unconfirmed for a ~55-person Series A SaaS company. This uncertainty pulls the holistic score slightly but not below 0.80. All eight claims are grounded with verbatim source quotes. With a holistic fit score of 0.85 and every claim grounded, the action threshold resolves to auto_add.

Chatmiss

  • 1/10 claim source quote(s) not in input
Action
propose ✓
Fit
0.88 (gold 0.78, |Δ| 0.10)
Industry
B2B SaaS ✓
Segment
Developer tooling / Observability infrastructure ✗
Seniority
VP ✓
Company size
51-200 ✓
Grounding
9/10 claims grounded (90.0%)
Claims (10)
  1. Tobi is VP Product at Stackveil, managing three PMs and owning the roadmap
    ✓ in input My day-to-day is split between roadmap, working closely with our three PMs
  2. Stackveil closed a Series A approximately eighteen months ago
    ✓ in input We closed our Series A about eighteen months ago
  3. Stackveil employs approximately 55 people
    ✓ in input Stackveil employs approximately 55 people across engineering, product, and go-to-market functions
  4. Stackveil builds distributed tracing and log aggregation infrastructure for backend engineering teams at mid-sized SaaS companies
    ✓ in input We build distributed tracing and log aggregation infrastructure for backend engineering teams at mid-sized SaaS companies
  5. Tobi is piloting an anomaly detection layer using ML that is currently in private beta with six design partners
    ✗ not in input Right now I'm piloting an anomaly detection layer that uses ML to surface meaningful signal from trace data. We're in a private beta with six design partners
  6. Tobi spent four years as a PM at Datadog and two years leading product at a smaller APM startup that was acqui-hired
    ✓ in input Before this, I was a PM at Datadog for four years, then spent two years leading product for a smaller APM startup that got acqui-hired
  7. Tobi's core problem statement is ensuring engineers trust what observability tools tell them
    ✓ in input I've been in developer tooling long enough to know that the hardest part isn't building the product — it's making sure engineers actually trust what it's telling them. That's the problem I came to Stackveil to work on.
  8. Customers are experiencing alert noise and false positives that need addressing
    ✓ in input spending a lot of time with customers who are drowning in alert noise and false positives
  9. The anomaly detection goal is to cut time-to-detection for P1 incidents without manual threshold tuning
    ✓ in input the goal is to cut time-to-detection for P1 incidents without requiring teams to hand-tune thresholds manually
  10. Tobi is based in London
    ✓ in input I'm based in London
Draft hook

Tobi — the framing you use about engineer trust resonated: anomaly detection that's technically sound but that on-call engineers second-guess at 2am is a solved problem that isn't solved. The tension between reducing false positives and actually surfacing signal that teams act on seems central to where you're taking the beta. Curious what patterns from your design partners have been most useful for calibrating that trust gap — and whether [relevant value prop] is worth exploring while you're still in that formative phase.

Reasoning

Tobi is a strong ICP fit: VP Product role at a Series A B2B SaaS company (55 people, likely $3M–$15M ARR) building developer tooling with an active ML feature in private beta. All five ICP dimensions align well. His stated priority — engineer trust in observability tooling — is concrete and defensible. He has deep domain expertise (Datadog, APM startup background) which means he will evaluate rigorously, but also that he's a high-value conversation for the right solution. The private beta with design partners creates a natural, low-friction conversation opening. Recommend `propose` as the action since this is a high-quality fit with a clear, personalized outreach angle that respects his sophistication and current moment in product development."