Software Engineering Intern

Job Description

About Company

Jingo is a stealth-mode startup focused on reimagining the future of shopping. They are developing a shopping OS that integrates intent, memory, and mood, aiming to replace traditional search with a more intuitive, personal, and emotionally aware experience. The company is backed by top-tier pre-seed investors.

Job Description

Job Title: Software Engineering Intern

Who can apply:
• Current enrolled college students
• Recent college graduates

Duration:
• 3 months, starting ASAP (no exceptions).
• Potential for extension beyond the initial 3 months.

Compensation:
• $6,000 per month.

Immigration:
• Not eligible for visa sponsorship (this includes CPT/OPT).

Location:
• In-person at Jingo’s SF office.
• Open to remote for the right candidate.

Portfolio Requirement:
• Applicants must include their GitHub profile or relevant project links.

Role Overview:
Jingo is seeking a Software Engineering Intern who thrives in fast-paced, high-ownership environments. This role is primarily backend-heavy and involves building LLM-powered infrastructure for:
• Natural-language search
• Real-time retrieval
• Contextual personalization

The intern will work across several key areas:
• Retrieval & ranking
• Preference modeling from implicit signals
• LLM tools integration and query transformation
• System scalability and reliability

Key Responsibilities:
• LLM & Retrieval Prototyping: Build prompts and pipelines to transform user input into structured intents; implement filtering and safety mechanisms.
• Vector & Metadata Search: Operate Pinecone (or equivalent) for managing embeddings, namespaces, and metadata filters.
• Backend APIs: Develop FastAPI services to power search, recommendation systems (recsys), and memory; define clean API contracts and internal SDKs.
• System Architecture: Design resilient and observable services incorporating caching (Redis), rate-limiting, timeouts, and fallbacks.
• Data Pipelines: Batch large-scale embedding and indexing tasks; manage concurrency, retries, and idempotency.
• Personalization Signals: Extract and store engagement signals (e.g., search queries, chat interactions, wishlist items, purchases) for use in downstream ranking.
• Performance & Reliability: Optimize low-latency APIs and high-throughput asynchronous pipelines.
• Collaboration: Work closely with the research team on retrieval evaluation and with product/design teams on user-facing results.

Qualifications (Required Skills & Experience):
• Strong Python proficiency (preferred), including:
• FastAPI
• `async/await` patterns
• Type-safe packaging
• Redis
• PostgreSQL
• REST/JSON principles
• Hands-on experience with LLM integration:
• Crafting prompts
• Handling tool outputs
• Managing JSON contracts
• Implementing fallback logic
• NLP foundations:
• Embeddings
• Vector search
• RAG (Retrieval Augmented Generation) systems
• Hybrid (keyword + metadata) search
• Experience building scalable APIs and async client wrappers for third-party APIs.
• Comfort with concepts in distributed systems, caching, observability, rate limiting, and backpressure.
• Familiarity with GCP (Google Cloud Platform), Docker, Git, and CI workflows.
• Clear communication skills, self-directed work ethic, and a strong sense of ownership.

Nice-to-Have / TS Path Eligible (Additional Beneficial Skills):
• Strong TypeScript / Node.js (especially NestJS) and a willingness to work across both Python and TypeScript stacks.
• Experience with Pinecone, Zilliz, or other vector databases (both for retrieval and upserting data).
• Recsys exposure: Preference modeling, ranking heuristics, and metadata schema design.
• Light frontend experience (React/Next.js) for wiring API touchpoints.
• Experience evaluating embedding models and understanding search quality tradeoffs.

Why Join Jingo:
• Impact: Opportunity to own core personalization systems from Day 1.
• Speed: Emphasis on building fast and iterating faster, valuing learning loops.
• Vision: Contribute to shaping an emotionally intelligent shopping experience powered by memory.
• Tools: Provided with a latest MacBook Pro and preferred setup.