- Geylang Southeast (Singapore) Singapore
全职 Full-Stack Engineer 工作, 薪水 up to SGD 5,000, CLIPLY PTE. LTD. Southeast (Singapore) 公司招聘中 - Ricebowl
分享
保存
工作地点
职位描述
岗位职责
Role Overview
We are hiring a Full‑Stack Engineer to build the first working version ofCliply’s product end‑to‑end. This role sits across backend, frontend, datamodelling and AI integration. You will work closely with our AI engineers toturn multimodal outputs (video perception, transcripts, reasoning, embeddings)into a usable product surface.
This is not a narrow backend orfrontend role. It is a founding full‑stack role for someone who has shippedcomplete web products independently and is comfortable owning both thetechnical and product behaviour.
What You Will Build (First 2–3Months)
1. Ingest Path
Register a licensed film or long‑form video.
Extract a playback proxy (lower‑resolution, streamable).
Maintain an exact timestamp mapping between source and proxy.
2. Subtitle Handling
Parse caption files (SRT, WebVTT).
Store every word with start/end timestamps in Postgres.
Maintain raw text separately from searchable text.
3. Provider Adapter
Send video to a managed video‑search or multimodal API.
Poll for readiness.
Convert provider output into Cliply’s internal candidate format.
4. Search Page
Query box with autocomplete.
List of search results with timestamps.
Video player that seeks to the exact moment.
Subtitle line that triggered the match shown underneath.
5. Correction Loop
Editor UI to adjust in/out points.
Accept/reject candidate results with reasons.
Persist corrections and editorial decisions.
Two Non‑Negotiables
Deletion propagation: removing an asset or revoking rights must remove it everywhere immediately.
Search correctness: a search that finds nothing must return nothing — never “closest match”.
Key Responsibilities
Full‑Stack Product Development
Build backend services using Python + FastAPI.
Build frontend interfaces using React + TypeScript.
Create clean, intuitive UI without a designer.
Own end‑to‑end product behaviour from ingest to correction.
Backend Architecture &Pipelines
Design REST/gRPC APIs for ingestion, retrieval, metadata access and multimodal outputs.
Build background workers and long‑running jobs with retries, idempotency and resumability.
Integrate slow, asynchronous or imperfect third‑party AI APIs gracefully.
Implement deletion propagation, rights revocation and consistency guarantees.
Data & Storage
Design Postgres schemas for subtitles, metadata, candidates and editorial corrections.
Build indexes for fast retrieval across multimodal outputs.
Store embeddings, timelines and metadata graphs efficiently.
Product Iteration
Sit with users, observe friction, and change the product rather than explain it.
Work closely with AI engineers to validate perception, alignment and reasoning outputs.
Deliver investor‑ready demos and internal prototypes.
Collaboration
Partner with AI engineers on multimodal pipeline integration.
Partner with backend engineers on scaling ingestion and orchestration.
Provide early frontend foundations until a dedicated frontend engineer joins.
Requirements
Must‑Have
You have built and shipped a complete web product end‑to‑end (backend, database, frontend, deployment).
Strong experience with Python (FastAPI) and React (TypeScript).
Deep comfort with Postgres: schema design, indexing, query optimisation.
Experience with long‑running background jobs and understanding of retries, idempotency and resumability.
Experience consuming slow, asynchronous or unreliable third‑party APIs.
Ability to build clean, obvious interfaces without a designer.
Strong product instincts: you iterate based on user behaviour, not assumptions.
Comfortable working in a fast‑moving pre‑seed environment.
Nice‑to‑Have
Experience with video processing (ffmpeg, HLS, proxies).
Experience with multimodal AI APIs (vision, speech, embeddings).
Experience with search systems (Elasticsearch, pgvector).
Experience with media or broadcast workflows.
Experience with cloud deployment (AWS/Azure/GCP).
We are hiring a Full‑Stack Engineer to build the first working version ofCliply’s product end‑to‑end. This role sits across backend, frontend, datamodelling and AI integration. You will work closely with our AI engineers toturn multimodal outputs (video perception, transcripts, reasoning, embeddings)into a usable product surface.
This is not a narrow backend orfrontend role. It is a founding full‑stack role for someone who has shippedcomplete web products independently and is comfortable owning both thetechnical and product behaviour.
What You Will Build (First 2–3Months)
1. Ingest Path
Register a licensed film or long‑form video.
Extract a playback proxy (lower‑resolution, streamable).
Maintain an exact timestamp mapping between source and proxy.
2. Subtitle Handling
Parse caption files (SRT, WebVTT).
Store every word with start/end timestamps in Postgres.
Maintain raw text separately from searchable text.
3. Provider Adapter
Send video to a managed video‑search or multimodal API.
Poll for readiness.
Convert provider output into Cliply’s internal candidate format.
4. Search Page
Query box with autocomplete.
List of search results with timestamps.
Video player that seeks to the exact moment.
Subtitle line that triggered the match shown underneath.
5. Correction Loop
Editor UI to adjust in/out points.
Accept/reject candidate results with reasons.
Persist corrections and editorial decisions.
Two Non‑Negotiables
Deletion propagation: removing an asset or revoking rights must remove it everywhere immediately.
Search correctness: a search that finds nothing must return nothing — never “closest match”.
Key Responsibilities
Full‑Stack Product Development
Build backend services using Python + FastAPI.
Build frontend interfaces using React + TypeScript.
Create clean, intuitive UI without a designer.
Own end‑to‑end product behaviour from ingest to correction.
Backend Architecture &Pipelines
Design REST/gRPC APIs for ingestion, retrieval, metadata access and multimodal outputs.
Build background workers and long‑running jobs with retries, idempotency and resumability.
Integrate slow, asynchronous or imperfect third‑party AI APIs gracefully.
Implement deletion propagation, rights revocation and consistency guarantees.
Data & Storage
Design Postgres schemas for subtitles, metadata, candidates and editorial corrections.
Build indexes for fast retrieval across multimodal outputs.
Store embeddings, timelines and metadata graphs efficiently.
Product Iteration
Sit with users, observe friction, and change the product rather than explain it.
Work closely with AI engineers to validate perception, alignment and reasoning outputs.
Deliver investor‑ready demos and internal prototypes.
Collaboration
Partner with AI engineers on multimodal pipeline integration.
Partner with backend engineers on scaling ingestion and orchestration.
Provide early frontend foundations until a dedicated frontend engineer joins.
Requirements
Must‑Have
You have built and shipped a complete web product end‑to‑end (backend, database, frontend, deployment).
Strong experience with Python (FastAPI) and React (TypeScript).
Deep comfort with Postgres: schema design, indexing, query optimisation.
Experience with long‑running background jobs and understanding of retries, idempotency and resumability.
Experience consuming slow, asynchronous or unreliable third‑party APIs.
Ability to build clean, obvious interfaces without a designer.
Strong product instincts: you iterate based on user behaviour, not assumptions.
Comfortable working in a fast‑moving pre‑seed environment.
Nice‑to‑Have
Experience with video processing (ffmpeg, HLS, proxies).
Experience with multimodal AI APIs (vision, speech, embeddings).
Experience with search systems (Elasticsearch, pgvector).
Experience with media or broadcast workflows.
Experience with cloud deployment (AWS/Azure/GCP).
重要安全守则
申请工作时，切勿提供您的银行或信用卡详细资料。不要转账或完成无关的在线调查问卷。如果您发现可疑内容，请举报此招聘广告。
了解更多