LLM & AI Engineering
Building product features on top of large language models.
Backend Engineer track
Junior
Write correct code, ship small changes safely, ask good questions.
- Large Language ModelA model trained on huge amounts of text to understand and generate language.
Mid-level
Own a feature end to end without hand-holding.
- AI AgentA model that plans and takes actions with tools in a loop.
- EmbeddingsNumeric vectors representing meaning, used for similarity search.
- How LLMs WorkTokens in, next-token prediction out: a mental model of what's happening.
- Multimodal ModelsModels that handle images, audio and text together.
- Prompt EngineeringWriting instructions that get reliable results from a model.
- Prompt InjectionUntrusted text hijacking a model's instructions.
- Retrieval-Augmented Generation (RAG)Giving a model relevant documents so it answers from your data.
- Semantic SearchSearching by meaning instead of keywords.
- Streaming ResponsesShowing model output token by token as it's generated.
- Structured OutputGetting models to return valid JSON matching a schema.
- System PromptInstructions that set a model's behavior for a conversation.
- TemperatureA setting controlling how random a model's output is.
- TokenThe unit of text a model reads and writes, and what you pay for.
- Tool Use / Function CallingLetting a model call your functions and APIs.
- Vector DatabaseStoring embeddings for similarity search.
Senior
Own a system, its failure modes, and its trade-offs.
- Choosing a ModelTrading off quality, speed and cost across model sizes and providers.
- ChunkingSplitting documents into pieces for embedding and retrieval.
- EvalsSystematically measuring the quality of LLM output.
- Fine-TuningFurther training a model on your own examples.
- GuardrailsChecks on model inputs and outputs for safety and correctness.
- LLM Cost and LatencyManaging tokens, model choice and caching.
- LLM-as-JudgeUsing one model to grade another model's output.
- Prompt CachingReusing processed prompt prefixes to save cost and time.
- RerankingRe-ordering retrieved results with a stronger model.
Staff
Shape how many teams build, across systems.
Nothing here yet.
Principal
Set technical direction for the organization.
Nothing here yet.
Data Analyst track
Junior
Write correct SQL, build trusted dashboards, ask good questions.
Nothing here yet.
Mid-level
Own an analysis end to end, from vague question to recommendation.
- How LLMs WorkTokens in, next-token prediction out: a mental model of what's happening.
- Prompt EngineeringWriting instructions that get reliable results from a model.
Senior
Own experimentation and metrics design; call out bad numbers.
Nothing here yet.
Staff
Shape how the organization measures and decides.
Nothing here yet.
Principal
Set measurement strategy across the company.
Nothing here yet.
Data Engineer track
Junior
Build and fix pipelines from clear specs; write correct SQL.
- Large Language ModelA model trained on huge amounts of text to understand and generate language.
Mid-level
Own pipelines and models end to end, including their quality.
- AI AgentA model that plans and takes actions with tools in a loop.
- EmbeddingsNumeric vectors representing meaning, used for similarity search.
- How LLMs WorkTokens in, next-token prediction out: a mental model of what's happening.
- Multimodal ModelsModels that handle images, audio and text together.
- Prompt EngineeringWriting instructions that get reliable results from a model.
- Prompt InjectionUntrusted text hijacking a model's instructions.
- Retrieval-Augmented Generation (RAG)Giving a model relevant documents so it answers from your data.
- Semantic SearchSearching by meaning instead of keywords.
- Streaming ResponsesShowing model output token by token as it's generated.
- Structured OutputGetting models to return valid JSON matching a schema.
- System PromptInstructions that set a model's behavior for a conversation.
- TemperatureA setting controlling how random a model's output is.
- TokenThe unit of text a model reads and writes, and what you pay for.
- Tool Use / Function CallingLetting a model call your functions and APIs.
- Vector DatabaseStoring embeddings for similarity search.
Senior
Design the platform's storage, processing and modeling choices.
- Choosing a ModelTrading off quality, speed and cost across model sizes and providers.
- ChunkingSplitting documents into pieces for embedding and retrieval.
- EvalsSystematically measuring the quality of LLM output.
- Fine-TuningFurther training a model on your own examples.
- GuardrailsChecks on model inputs and outputs for safety and correctness.
- LLM Cost and LatencyManaging tokens, model choice and caching.
- LLM-as-JudgeUsing one model to grade another model's output.
- Prompt CachingReusing processed prompt prefixes to save cost and time.
- RerankingRe-ordering retrieved results with a stronger model.
Staff
Shape how the whole organization produces and uses data.
Nothing here yet.
Principal
Set data strategy and architecture across the company.
Nothing here yet.
Frontend Engineer track
Junior
Build UI that works, ship small changes safely, ask good questions.
- Large Language ModelA model trained on huge amounts of text to understand and generate language.
Mid-level
Own a feature end to end without hand-holding.
- AI AgentA model that plans and takes actions with tools in a loop.
- EmbeddingsNumeric vectors representing meaning, used for similarity search.
- How LLMs WorkTokens in, next-token prediction out: a mental model of what's happening.
- Multimodal ModelsModels that handle images, audio and text together.
- Prompt EngineeringWriting instructions that get reliable results from a model.
- Prompt InjectionUntrusted text hijacking a model's instructions.
- Retrieval-Augmented Generation (RAG)Giving a model relevant documents so it answers from your data.
- Semantic SearchSearching by meaning instead of keywords.
- Streaming ResponsesShowing model output token by token as it's generated.
- Structured OutputGetting models to return valid JSON matching a schema.
- System PromptInstructions that set a model's behavior for a conversation.
- TemperatureA setting controlling how random a model's output is.
- TokenThe unit of text a model reads and writes, and what you pay for.
- Tool Use / Function CallingLetting a model call your functions and APIs.
- Vector DatabaseStoring embeddings for similarity search.
Senior
Own an app's architecture, performance, and failure modes.
- AI Product UXDesigning interfaces for uncertain, streaming, sometimes-wrong AI output.
- Choosing a ModelTrading off quality, speed and cost across model sizes and providers.
- EvalsSystematically measuring the quality of LLM output.
- Fine-TuningFurther training a model on your own examples.
- GuardrailsChecks on model inputs and outputs for safety and correctness.
- LLM Cost and LatencyManaging tokens, model choice and caching.
- LLM-as-JudgeUsing one model to grade another model's output.
Staff
Shape how many teams build, across apps.
Nothing here yet.
Principal
Set technical direction for the organization.
Nothing here yet.