AI Jobs Map

LH2 AI Labs · Bengaluru, Karnataka, India

Founder's Office - Sprint Team Lead

executivefull timePosted 15 days ago
Apply on LinkedInOpens the original posting. AI Jobs Map never asks for your details.

Stack mentioned

agentic-ai

Sprint Team Lead

Founder's Office – Rapid Sourcing & Fulfillment, Multimodal

LH2 AI Labs · 2 - 5 yrs experience · On-site

ABOUT LH2

Built by second-time founders who have built and sold companies before, LH2 AI Labs (lh2.ai) is building the post-training infrastructure for frontier AI models. We bring private, high-quality institutional datasets and vetted domain experts to frontier AI pipelines across verticals — coding, computer use, agentic, medical, audio, and more.

For AI to keep progressing, it needs high-quality datasets drawn from real production use cases. The public web has already been crawled and trained on — there's nothing left in the public domain to learn from. That's where we come in: we partner with institutions to connect their data to frontier models and help advance AGI.

Not every request fits neatly into one of our named verticals. Every week, clients — frontier labs and the data marketplaces that feed them — come to us with one-off, time-boxed asks spanning many kinds of data: audio, video, imaging, documents, structured data, and more. The Sprint Team exists to catch every one of these, scope it quickly, and go find who in the world can supply it — often on a tight timeline. We're hiring the person who runs that team.

THE ROLE
We're looking for someone to build and run our Sprint Team.

This is a different shape of the founder's-office seat than our vertical-lead roles. Where a Vertical Lead goes deep on one data category - Healthcare, Company Ops, Code - for months at a time, the Sprint Team goes wide. The modality, format, and timeline vary from one request to the next. Your job is to take whatever lands in the queue, scope it clearly, and mobilize your team to go find it.

You'll run this with a small team under you - GTM Analysts and Interns - and it's worth being upfront about what that means: the team gives you leverage, but it isn't a step back from the work itself. You'll stay close to requests yourself — sourcing vendors, getting on calls, following up on samples - while the team lets more move in parallel. It's a good fit if you want to keep building hands-on rather than just assigning and reviewing.

Requests come in with their own deadlines, and you'll often be working a few at once across different kinds of data. You'll be judged on how much you get closed, and how reliably.

We can teach you any given modality - audio specs, imaging formats, licensing terms. What matters more is staying organized and steady when there's a lot happening at once. That part you bring.

WHAT YOU'LL ACTUALLY DO
Run the Queue

- Triage every incoming request the day it lands: pin down exactly what the client needs, by when, and what “good” looks like — the tracker only works if the requirement brief is real, not a placeholder.

- Turn requests that show up with almost no detail into something your team can act on: format, volume, technical spec, language, sample expectations.

- Personally run point on the hardest or most time-sensitive requests — the ones with a 2–3 day sample deadline or a modality nobody on the team has sourced before.

- Keep the vendor tracker honest: every vendor contacted, every sample requested, every price quoted, logged and current — not reconstructed after the fact.

Build & Catalogue the Vendor Database

- Build out a structured, searchable database of vendors across data modalities — audio, video, imaging, documents, structured data — so sourcing starts from a warm list, not from zero every time.

- Capture what matters for each vendor: capability, geography, capacity, pricing, quality track record, and past samples — kept current, not reconstructed after the fact.

- Continuously expand and refine the database as new requests surface, new vendors and new categories of supply.

- Make it genuinely usable by your team — clear tagging and categorization so analysts and interns can find the right vendor fast, without needing you in the loop.

Build & Lead the Team

- Hire, ramp, and manage your GTM Analysts and Interns — split the queue so nothing sits unworked, and know at all times who owns what.

- Turn cold-outreach and vendor-vetting into a playbook your analysts and interns can run without you in the room: how to find a vendor for a modality you've never sourced before, how to qualify them fast, how to get a sample moving.

- Review what your team sources before it goes to the client — you're the last quality gate on volume, pricing, licensing terms, and sample fit.

- Coach analysts and interns on negotiation and objection-handling as they take on more of the closing themselves.

Own the Function

- Build the intake-to-close process for one-off requests from scratch, and keep improving it as a system that scales past you.

- Report weekly to the founders on what came in, what closed, what's stuck, and where the team needs more supply relationships before the next urgent task hits.

WHAT WE'RE LOOKING FOR

- Genuinely sharp. Top-tier education and/or time at well-regarded companies — you've operated around a high bar and it shows. You pick up an unfamiliar data category fast enough to sound credible on a call about it within a day.

- Breadth over depth. You'd rather solve five unrelated problems this week than spend the year becoming the world's expert in one.

- Real urgency. A three-day sample deadline doesn't scare you off — it's the job. You default to moving now, not to a plan for moving later.

- Some experience leading or mentoring others, even informally — you'll have a team looking to you to unblock them every day.

- You can sell and you can source cold. 3–6 years in sales, partnerships, BD, or vendor/supply sourcing with a real track record of opening relationships from scratch — not just managing an existing book.

- Process instinct — you notice when a shared tracker or workflow is breaking down and you fix it without being asked.

- Comfortable being the person who says “we don't have this yet” and then goes and gets it, across data types you've never touched before.

- High ownership; thrive in fast, ambiguous environments.

More jobs at LH2 AI Labs