Item 23 · synthetic
Synthetic Ambig Role Adjacent Staff Eng
scenario synthetic_ambig_role_adjacent_staff_eng
Input
Declan Brennan Staff Engineer at Vaultry (Series A, ~55 people) I joined Vaultry about two years ago as one of the first engineers on the platform team, and since then I've been heads-down building out the core infrastructure that sits underneath our cloud security posture product. Most of my work lives at the intersection of data ingestion and detection logic — we pull signals from AWS, GCP, and Azure configurations, normalize them into a unified graph model, and run a set of policy rules that surface misconfigurations to security teams before they become incidents. More recently I've been a significant contributor to the ML feature we shipped earlier this year, which clusters configuration drift patterns to surface anomalies that static rules miss. I wrote the initial feature pipeline and spent a lot of time with our two data scientists getting the training cadence right. It's in production now with a handful of design-partner customers, and the early feedback has been encouraging. Before Vaultry I spent four years at a larger infrastructure software company in Dublin, working on distributed systems for network monitoring. That background has been useful here — the scale and latency constraints are different, but the instinct for reliability carries over. I don't have a formal remit over product direction or roadmap, but I work closely with the product team and I'm usually in the room when technical feasibility decisions get made. Dublin-based, working mostly within the European enterprise market. Vaultry is a Dublin-based B2B SaaS company providing cloud security posture management tooling for mid-market and enterprise customers. Founded in 2020, the company raised a Series A round in 2022 and currently employs around 55 people across engineering, product, and go-to-market functions. Vaultry's platform continuously monitors cloud infrastructure configurations across major providers, identifying misconfigurations and policy violations in real time. The product includes an anomaly detection layer built on ML that clusters configuration drift patterns to surface risks static rule engines miss. Customers are primarily security and DevOps teams in financial services and SaaS verticals across Europe.
Gold
- Expected action
- propose
- Fit score
- 0.72
- Industry
- B2B SaaS
- Segment
- Cloud security posture management
- Seniority
- IC
- Company size
- 51-200
- Language
- en
- Notes
- Staff Engineer at small Series A B2B SaaS — rubric 0.75 adjacent anchor. Product and stage clean; arr just under floor.
Predictions
Integratedmiss
- action discard, expected propose
- Action
- discard ✗ (gold: propose)
- Fit
- 0.45 (gold 0.72, |Δ| 0.27)
- Industry
- B2B SaaS ✓
- Segment
- Cloud Security Posture Management (CSPM) ✓
- Seniority
- IC ✓
- Company size
- 51-200 ✓
- Grounding
- 6/6 claims grounded (100.0%)
Claims (6)
- Vaultry is a Series A B2B SaaS company.✓ in input
Series A, ~55 people - Vaultry employs approximately 55 people.✓ in input
currently employs around 55 people across engineering, product, and go-to-market functions - Vaultry has shipped a production ML feature that clusters configuration drift patterns to surface anomalies.✓ in input
ML feature we shipped earlier this year, which clusters configuration drift patterns to surface anomalies that static rules miss - The ML feature is live with early customers and receiving positive feedback.✓ in input
It's in production now with a handful of design-partner customers, and the early feedback has been encouraging - Declan Brennan is a Staff Engineer — an individual contributor role with no formal product or roadmap remit.✓ in input
I don't have a formal remit over product direction or roadmap - Declan works closely with the product team and is involved in technical feasibility decisions.✓ in input
I work closely with the product team and I'm usually in the room when technical feasibility decisions get made
Draft hook
Hi Declan — saw that Vaultry shipped an ML layer earlier this year that clusters configuration drift patterns to surface anomalies static rules miss. Impressive to get that into production with design-partner customers already. Given how closely you work with the product team on technical feasibility decisions, I thought it worth a quick note — [product pitch]. Happy to keep it brief.
Reasoning
The company (Vaultry) is a strong fit on nearly every company-level dimension: confirmed Series A stage (1.0), headcount of ~55 squarely within the 20–250 range (1.0), and a B2B SaaS product with a shipped, production AI/ML feature (1.0). ARR is unattested in the input, so that dimension scores low (0.3) but is not sufficient alone to trigger a refuse — there is enough signal to render an overall judgment. The decisive failure is role match (0.1). Declan Brennan is a Staff Engineer — an explicitly IC-level individual contributor. The ICP targets VP Product, Head of AI/ML, Director of Engineering, Founder, or CTO. He himself states he has no formal remit over product direction or roadmap. While he is involved informally in technical feasibility discussions, he lacks the authority, budget ownership, or strategic mandate that defines the ICP persona. Outreach to him is unlikely to generate a qualified opportunity directly. All six claims are grounded in verbatim quotes, so the action is determined purely by the sub-0.50 holistic fit score → discard.
Chatpass
- Action
- propose ✓
- Fit
- 0.68 (gold 0.72, |Δ| 0.04)
- Industry
- B2B SaaS ✓
- Segment
- Cloud security posture management ✓
- Seniority
- IC ✓
- Company size
- 51-200 ✓
- Grounding
- 8/8 claims grounded (100.0%)
Claims (8)
- Declan was the primary builder of the ML anomaly detection feature shipped earlier this year✓ in input
I wrote the initial feature pipeline and spent a lot of time with our two data scientists getting the training cadence right. It's in production now with a handful of design-partner customers - Declan participates in technical feasibility decisions alongside the product team✓ in input
I work closely with the product team and I'm usually in the room when technical feasibility decisions get made - Declan does not have formal authority over product direction or roadmap✓ in input
I don't have a formal remit over product direction or roadmap - Vaultry is Series A stage with ~55 employees✓ in input
Series A, ~55 people - Vaultry's product monitors cloud configurations across AWS, GCP, and Azure and surfaces misconfigurations✓ in input
we pull signals from AWS, GCP, and Azure configurations, normalize them into a unified graph model, and run a set of policy rules that surface misconfigurations to security teams before they become incidents - The ML feature clusters configuration drift patterns to surface anomalies that static rules miss✓ in input
the ML feature we shipped earlier this year, which clusters configuration drift patterns to surface anomalies that static rules miss - Declan has prior experience with distributed systems and network monitoring infrastructure✓ in input
I spent four years at a larger infrastructure software company in Dublin, working on distributed systems for network monitoring - Vaultry focuses on European enterprise market✓ in input
Dublin-based, working mostly within the European enterprise market
Draft hook
Saw that Vaultry shipped an anomaly detection layer this year clustering configuration drift patterns — that's a harder problem than it looks, especially keeping training cadence stable on infrastructure signals that shift constantly. Curious what you've learned running it with design partners. We work with a few teams solving similar problems at the feature pipeline layer and I wondered if any of it would be relevant to where you're taking it next.