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

← Scorecard

Item 69 · edge

Edge Ambiguous Title Technical Lead

scenario edge_ambiguous_title_technical_lead

Input

Dmitri Petrov
Technical Lead at Stackform (Series A, ~50 people)

I joined Stackform about two and a half years ago when the engineering team was still small enough to fit around one table. My day-to-day splits between hands-on architecture work and keeping four engineers moving forward without friction — code reviews, design conversations, unblocking the occasional Kubernetes rabbit hole.

On the technical side, I spend most of my time on our infrastructure abstraction layer: the piece that lets customers define cloud resources in a declarative way and have Stackform reconcile the desired state against AWS, GCP, or Azure. We made an early call to build on top of a custom operator pattern rather than adapting an existing framework, and I still think that was the right trade-off for the level of configurability our customers need.

We have an AI-assisted drift detection feature in active development right now. The idea is to surface anomalies in infrastructure state before they become incidents, using a model trained on customers' own historical telemetry. I'm leading the integration work on that — figuring out where the model outputs slot into our existing alerting pipeline and what the latency envelope needs to look like to be genuinely useful.

Before Stackform I worked at a managed hosting company and a smaller consultancy focused on cloud migrations. I tend to prefer systems that fail clearly over systems that fail quietly.

Stackform is a B2B SaaS company building a cloud infrastructure management platform for engineering teams running workloads across multiple cloud providers. Founded in 2021 and backed by a Series A round, the company employs around fifty people and serves mid-market customers who need consistent, auditable control over their cloud environments without maintaining large platform teams. Stackform's product offers declarative infrastructure configuration, real-time state reconciliation across AWS, GCP, and Azure, and role-based policy enforcement. The company is currently developing an AI-assisted drift detection capability designed to identify anomalies in infrastructure state using customer-specific telemetry, with the goal of reducing mean time to detection for configuration-related incidents.

Gold

Expected action
propose
Fit score
0.65
Industry
B2B SaaS
Segment
DevOps / cloud infrastructure management
Seniority
—
Company size
11-50
Language
en
Notes
Ambiguous title. Bio is consistent with Staff Engineer / Tech Lead with mentoring duties — neither IC-without-authority nor confirmed manager. Held role_match at 0.5.

Predictions

Integratedmiss

  • classification miss: segment, seniority
Action
propose ✓
Fit
0.62 (gold 0.65, |Δ| 0.03)
Industry
B2B SaaS ✓
Segment
Cloud infrastructure management / multi-cloud platform engineering ✗
Seniority
IC ✗
Company size
11-50 ✓
Grounding
6/6 claims grounded (100.0%)
Claims (6)
  1. Stackform is a B2B SaaS company.
    ✓ in input Stackform is a B2B SaaS company building a cloud infrastructure management platform for engineering teams running workloads across multiple cloud providers.
  2. Stackform is Series A-backed.
    ✓ in input backed by a Series A round
  3. Stackform employs around fifty people.
    ✓ in input employs around fifty people
  4. Stackform has an AI-assisted drift detection capability in active development.
    ✓ in input currently developing an AI-assisted drift detection capability designed to identify anomalies in infrastructure state using customer-specific telemetry
  5. Dmitri is leading the AI integration work for the drift detection feature.
    ✓ in input I'm leading the integration work on that — figuring out where the model outputs slot into our existing alerting pipeline and what the latency envelope needs to look like to be genuinely useful.
  6. The drift detection feature uses a model trained on customers' own historical telemetry.
    ✓ in input using a model trained on customers' own historical telemetry
Draft hook

Hi Dmitri — I noticed Stackform is currently developing an AI-assisted drift detection capability using customer-specific telemetry, and that you're leading the integration work on slotting model outputs into your alerting pipeline. We work with B2B SaaS teams at exactly this stage — taking a model from development into a production alerting path — and the latency and reliability constraints you're navigating are ones we've seen come up repeatedly. Worth a 20-minute conversation to compare notes?

Reasoning

**Company fit is strong; role fit is the main drag.** - **Stage (1.0):** Explicitly Series A — a direct ICP match. - **Headcount (0.90):** "Around fifty people" sits squarely inside the 20–250 range, with a small rounding-uncertainty discount. - **ARR (0.50):** No ARR figure is provided anywhere in the input. A Series A B2B SaaS with ~50 employees founded in 2021 is *plausibly* in the $2M–$50M range, but there is no verbatim evidence to move this above neutral. - **Product shape (1.0):** Both the profile and company text explicitly confirm an AI-assisted drift detection feature "in active development," satisfying the ICP criterion for a B2B SaaS with a user-facing AI feature in active development. - **Role (0.35):** Dmitri's title is **Technical Lead** — an IC/senior-IC role. The ICP targets VP Product, Head of AI/ML, Director of Engineering, or a technical Founder/CTO. He is doing relevant AI integration work and manages four engineers, but he is not a named decision-maker or budget holder per the ICP definition. This is the most significant misalignment. **Holistic score: 0.62** — the company is an excellent fit on most structural dimensions, but the contact is below the seniority threshold. This warrants a `propose` (human review) rather than auto-add, both because the score is in the [0.50, 0.80] band and because a senior stakeholder (Head of AI, CTO, VP Engineering) should likely be the primary outreach target. Dmitri could still be a valuable warm introduction path given his direct ownership of the AI feature.

Chatmiss

  • classification miss: segment, seniority, company_size
Action
propose ✓
Fit
0.75 (gold 0.65, |Δ| 0.10)
Industry
B2B SaaS ✓
Segment
Infrastructure/DevOps ✗
Seniority
Manager ✗
Company size
51-200 ✗
Grounding
7/7 claims grounded (100.0%)
Claims (7)
  1. Dmitri is a hands-on technical lead who straddles architecture and engineering management
    ✓ in input My day-to-day splits between hands-on architecture work and keeping four engineers moving forward without friction
  2. He owns the infrastructure abstraction layer, a core piece of Stackform's product
    ✓ in input On the technical side, I spend most of my time on our infrastructure abstraction layer: the piece that lets customers define cloud resources in a declarative way and have Stackform reconcile the desired state against AWS, GCP, or Azure.
  3. He is actively leading the integration work on the AI-assisted drift detection feature
    ✓ in input I'm leading the integration work on that — figuring out where the model outputs slot into our existing alerting pipeline and what the latency envelope needs to look like to be genuinely useful.
  4. Stackform is Series A with ~50 employees
    ✓ in input Founded in 2021 and backed by a Series A round, the company employs around fifty people
  5. The company is actively developing an AI-assisted drift detection feature using customer-specific telemetry
    ✗ not in input the company is currently developing an AI-assisted drift detection capability designed to identify anomalies in infrastructure state using customer-specific telemetry
  6. Dmitri's technical constraint involves mapping model outputs to actionable alerts while maintaining latency performance
    ✓ in input figuring out where the model outputs slot into our existing alerting pipeline and what the latency envelope needs to look like to be genuinely useful
  7. He values systems that fail clearly over systems that fail quietly
    ✓ in input I tend to prefer systems that fail clearly over systems that fail quietly.
Draft hook

Saw that Stackform is building AI-assisted drift detection on top of customer telemetry—and that you're leading the integration side. The latency-vs-usefulness tradeoff in that alerting pipeline is genuinely tricky, especially when model confidence scores need to map cleanly to actionable alerts rather than noise. We work with teams at a similar stage on exactly that problem. Worth a 20-minute conversation to see if there's anything useful to compare notes on?

Reasoning

Stackform is a strong company-level ICP fit (Series A, 50 people, B2B SaaS, active AI feature development). Dmitri is the right *technical* contact—he's directly owning the AI drift detection integration and wrestling with real constraints (latency, model-to-alert mapping). However, he's a Technical Lead, not the target decision-maker role (VP/Director/CTO). He's best positioned as an influencer or internal champion who would flag this to the economic buyer. The proposal action reflects high product/initiative fit but acknowledges that closing would likely require elevation to the CTO or VP Engineering. The hook is specific to his stated technical problem and uses a low-friction "compare notes" framing appropriate for a technical IC/manager hybrid."