Product Manager

Company Overview

Lifesight is a fast-growing SaaS company focused on helping businesses leverage data & AI to improve customer acquisition and retention. We have a team of 130 serving 300+ customers across 5 offices in the US, Singapore, India, Australia, and the UK.

Our mission is to make it easy for non-technical marketers to leverage advanced data activation and marketing measurement tools that are powered by AI, to improve their performance and achieve their KPIs. Our product is being adopted rapidly globally and we need the best people onboard the team to accelerate our growth.

Role Summary

You'll be the product lead in a compact, high‑output pod (Senior Fullstack AI Engineer, Junior Fullstack AI Engineer, Marketer, Ops). You will shape the problem, define the value hypothesis, and
co‑build

prototypes yourself (prompting, lightweight coding, RAG scaffolding) to de‑risk early, then drive an MVP to production in weeks. You're equal parts PM, experiment designer, and AI tinkerer.

What You'll Do

Discovery & Strategy

• Frame problems, jobs‑to‑be‑done, and value hypotheses; define target user, must‑win use‑cases, and success metrics.
• Map incumbents/workarounds; identify where AI gives a 10× edge (personalization, automation, prediction, UX).
• Prioritize ruthlessly for a 6‑week MVP; maintain a lean roadmap and cut scope aggressively to hit outcomes.

Hands‑On AI Prototyping

• Build quick experiments yourself (prompt/system‑instruction design, function/tool calling, structured outputs).
• Stand up lightweight scripts/notebooks or small services to test feasibility (Python/TypeScript + LangChain/LlamaIndex).
• Assemble RAG spikes: data scoping, chunking strategy, embeddings, vector search, context windows, eval sets.
• Create prompt/eval repos; run side‑by‑side model tests; instrument quality/latency/cost telemetry.

Product Execution

• Write crisp specs: user stories, acceptance criteria, data contracts, analytics events, safety/guardrail requirements.
• Drive design with low‑fi wireframes and clickable prototypes (Figma); specify UX for streaming and error states.
• Partner with engineers to sequence delivery; unblock daily; manage risk via spikes, flags, and phased rollouts.
• Ensure observability from day 1 (events taxonomy, funnels, cohorting); define experiment plans (A/Bs) and launch checklists.

AI Quality, Safety & Compliance

• Define guardrails and evaluation protocols (red‑team prompts, refusal/citation policies, safety filters, fallbacks).
• Track AI quality signals (accuracy, coverage, helpfulness, CSAT); iterate prompts/models and retrieval strategy.
• Coordinate privacy/security reviews; ensure PII handling, data retention, and vendor usage meet studio standards.

Go‑to‑Market & Outcomes

• Craft positioning and messaging with the marketer; align activation/onboarding to the AI value moments.
• Set adoption/engagement targets (activation %, WAU, task success, LLM cost per task) and run weekly metric reviews.
• Close the loop with qualitative feedback (surveys, interviews, support/chat logs) to inform the next iteration.

Minimum Qualifications

• 4–8 years in product management delivering 0→1 or growth iterations; startup or studio experience preferred.
• Hands‑on

with AI: can design prompts/system instructions, wire up function/tool calling, and build quick RAG spikes.
- Working proficiency in
Python. Nextjs or TypeScript

for prototyping; comfortable with APIs/SDKs and basic CLI/git.
- Strong fluency with product analytics (PostHog/Mixpanel/Amplitude), events taxonomies, and experiment design.
- Experience specifying and shipping AI‑powered UX (streaming, uncertainty, fallbacks, human‑in‑the‑loop).
- Excellent written specs, prioritization, and cross‑functional facilitation; bias to ship and learn.

Nice to Have

• Prior experience in SaaS creation tools, or workflow automation products.
• Knowledge of vector search (pgvector/Pinecone/Weaviate), embeddings, and retrieval strategies.
• Prompt/model evaluation tooling (Langfuse/Humanloop), LLM observability (Helicone), and OpenTelemetry.
• SQL for self‑serve analysis; dashboarding (Metabase/Looker); familiarity with data privacy frameworks.
• Light design skills (Figma) and comfort running user interviews and usability tests.