Miju Labs

← All buyers · Google DeepMind

H

Hannah Kirkwood

Human Data Engagements Lead, Google DeepMind · Google DeepMind

low confidenceTier A · Talk to firstFrontier labstext

Kirkwood leads 'human data engagements' at Google DeepMind out of the UK - on the face of the title, the person who sets up and runs GDM's engagements with the external rater pools and data suppliers that feed Gemini and the generative-media models. She has essentially no public footprint beyond her LinkedIn headline, so everything below her title is inference from how GDM describes its data-enrichment work.

Background

No verifiable public career history was found. The LinkedIn profile is UK-based, consistent with GDM's London operation, where the responsibility, ethics and data-enrichment functions sit. GDM's public posture on this work dates to its 2022 'Best practices for data enrichment' post (Antonia Paterson and Will Hawkins), co-developed with Partnership on AI, which commits to paying data workers above local living wage, piloting tasks before scale-up, verifying instructions and keeping an open channel to workers; GDM also runs a Human Behavioural Research Ethics Committee (HuBREC) modelled on an academic IRB. An engagements lead is the operational counterpart of that policy.

What they run now

Inferred: standing up and managing the supplier and rater-pool side of GDM human data - scoping engagements, onboarding platforms, holding them to the data-enrichment best practices, and routing capacity to the teams that need it. GDM's published rating volumes are large: the Imagen 3 report describes 366,569 ratings collected from 3,225 raters in 71 countries through an external platform, and the Veo and Nano Banana launches lean on human side-by-side comparisons for subjective visual quality. Someone in an engagements role is in the path of that spend.

What they buy

Inference, not evidence. Plausibly the commercial and operational counterparty for rater-pool capacity and annotation platforms across Gemini and generative media: side-by-side preference collection on images and video, rubric-driven quality rating, multilingual rater sourcing, and the compliance wrapper (living-wage pay, worker wellbeing, instruction verification) that GDM has publicly committed to. No vendor relationship of hers is documented publicly.

Career

  1. presentHuman Data Engagements Lead, Google DeepMind — Title from LinkedIn headline; start date not established.
Why it matters here

Potentially the highest-value name in this batch for Julian. GDM is the largest publicly evidenced buyer of subjective visual-quality judgement - the Imagen 3 numbers alone are a creative-taste data programme at industrial scale - and an 'engagements lead' title is exactly the function a specialist vendor has to get through. The caveat is honest: the title is all the evidence there is, and it could equally describe an internal programme role with no procurement authority.

How to reach

Speculative. She has no public speaking or writing to reference, so a cold approach has to lead with substance rather than flattery: a short note tying a creative/taste-data offer to GDM's own published commitments (the data-enrichment best practices and the Imagen 3 rater methodology), and proposing a small paid pilot on image or video side-by-sides with an expert rater panel. Do not open with a generic annotation pitch - GDM has that covered.

What we could not establish
  • Which team inside GDM she sits in (Gemini Data, Responsibility, or a central human-data function).
  • Whether 'engagements' means external supplier relationships or internal stakeholder programmes.
  • Her seniority and whether she holds budget.
  • Career history before GDM and how long she has been in the role.

Why they are on the map

LinkedIn profile: "Senior Program Manager (Human Data)" at Google DeepMind, London, England, United Kingdom, in role since 2022-09. source — evidence dated 2026-09

Runs vendor projects; decides if a pilot converts.

How Google DeepMind buys data

Imagen 3 report (2024-08) collected 366,569 ratings from 3,225 raters in 71 countries via an unnamed 'external platform' under GDM data-enrichment policy (living wage). Veo 3.1 page cites human side-by-sides on MovieGenBench (1,003 prompts), VBench I2V and internal sets. Nano Banana leads (2025) describe large-scale human evals plus trusted testers and artist reviews; Gemini Data team job post (2026) covers data curation/generation/evaluation.

Aug 2026: Kavukcuoglu runs GDM incl. Gemini as SVP; Hassabis chair. Gemini has Data, Post-Training, Safety Post-Training and Model Behavior teams (job posts); Gemini app sits inside GDM under Josh Woodward. Gen-media (Imagen/Nano Banana/Veo/Flow) has product lead Brichtova and modality research leads (Wang, Erhan).

Sources

  1. Hannah Kirkwood - Human Data Engagements Lead @ GDM (LinkedIn)
  2. Best practices for data enrichment - Google DeepMind