Company details

HEADCOUNT

25-99

ORG TYPE

Startup

SECTOR

Technology & Digital

WEBSITE

https://www.roofline.ai

About the company

Roofline builds an AI deployment stack for edge devices, centered on a retargetable AI compiler built on MLIR and IREE. The company positions the product as “one Python call” deployment across diverse SoCs, with a runtime that supports heterogeneous execution across CPU, GPU, and NPU. Public materials focus on making modern GenAI and vision models practical on constrained hardware. Hiring content is heavily engineering-led, especially around compilers and low-level runtime work.

Locations and presence

Roofline lists Cologne as headquarters, with roles posted in Köln and company legal details pointing to an Aachen address. The company’s footprint appears primarily Germany-based, with hybrid options stated on some roles.

Palpable Score

49.9

/ 100

Roofline is meaningfully accessible for early-career technical candidates through a real mix of intern, Master’s thesis, and junior full-time compiler roles. The score is held back by missing public detail on interview structure, onboarding and mentoring, pay ranges, and early-career progression outcomes.

Pillar 1: Early-career access

Score

14.0

/ 20

The company advertises student entry points including “Intern Software Engineer – AI Compilers” and a “Master Thesis” track with listed thesis topic areas.
Roofline posts full-time junior options like “Software Engineer for Host Compiler and Runtime (Junior)” alongside a combined “AI Compiler Engineer (Junior/Senior)” role.
The company keeps early-career access focused on deep systems work, which is strong for the right candidates but narrow for non-compiler entry-level paths.

Pillar 2: Hiring fairness and transparency

Score

11.8

/ 20

The company uses a structured application form and explicitly provides an application support email for upload issues, which is more reliable than “DM us” hiring.
Roofline publishes candidate data handling specifics in the recruitment privacy statement, including who can access applications and how long data is retained.
The company does not publicly outline interview stages, expected timelines, or assessment formats, so early-career candidates can’t judge process load up front.

Pillar 3: Learning and support

Score

8.8

/ 20

The company frames roles around working “alongside” other engineers and explicitly calls out close collaboration (for example in the AI Compilers internship mission section).
Roofline promises internal culture investment via regular team events and retreats, which can support learning loops if paired with real coaching.
The company does not describe onboarding plans, mentorship, manager 1:1 cadence, or feedback cycles in job pages, so support quality for first-job hires is hard to verify.

Pillar 4: Pay fairness and stability

Score

9.5

/ 20

The company states that equity is offered to every employee, including on junior and intern-facing postings, which is a fairness positive if consistently applied.
Roofline markets flexibility around schedules and time off, which can help early-career hires sustain high-intensity systems work.
The company does not publish salary ranges or benefits detail (beyond equity and flexibility language), which caps confidence on pay fairness.

Pillar 5: Early-career outcomes

Score

5.8

/ 20

The company does not publish early-career outcome evidence such as internship-to-full-time conversion rates, promotion examples, or levelling.
Roofline has limited third-party employee and interview feedback available publicly that is clearly attributable to the same Roofline entity and role set.
The company’s public team footprint indicates active hiring and growth, but retention and progression for juniors can’t be confirmed from public materials.