echoloc
Get started
echoloc/ Companies/ Perplexity

Perplexity Tech Stack

AI answer engine powered by search and large language models

Software Development San Francisco, California 201–500 employees Founded 2022 Privately Held
perplexity.ai LinkedIn

Perplexity operates an answer engine combining search, retrieval-augmented generation (RAG), and large-scale LLM inference. The tech stack reveals a production-heavy orientation: Kafka, Flink, Spark, and Dagster power real-time data pipelines; Kubernetes and Terraform manage infrastructure at scale; PostgreSQL, DynamoDB, Cassandra, and ClickHouse support varied query patterns. Active adoption of RAG and dbt signals maturation of grounding and data transformation layers, while pain points cluster around inference bottlenecks, search quality, and cluster utilization—classic scaling challenges for inference-driven products.

Tech Stack 124 technologies

Core StackFigma JavaScript Adobe Creative Cloud After Effects AWS Kubernetes Terraform Python Go Rust PostgreSQL MySQL DynamoDB Cassandra TypeScript React Databricks Snowflake Apache Spark Kafka Apache Flink Apache Airflow Dagster dbt Iceberg Delta Lake ClickHouse VPC AWS Transit Gateway Kinesis+92 more
AdoptingTerraform dbt RAG
ReplacingFramer

What Perplexity Is Building

◆Challenges

  • Improving search quality
  • Bottlenecks in inference stack
  • Maintaining high uptime
  • Scalable llm deployment
  • Improving ci/cd pipelines
  • Scaling data team capacity
  • Replacing repetitive tasks with self-service
  • Optimizing cluster utilization
  • Diagnosing performance bottlenecks
  • Improving workplace services

▲Active Projects

  • Search platform and model stack components
  • Develop apis for ai inference for internal and external customers
  • Large-scale deployment of machine learning models for real-time inference
  • Agent-driven cloud infrastructure management
  • Production workflow and intake process
  • Training pipelines and inference services apis
  • Develop llm-as-a-judge systems
  • Ai-readable data warehouse
  • Ai/ml pipeline security assessment
  • Rag pipelines for grounding and answer generation

Hiring Activity

Accelerating75 roles · 30 in 30d

Department

Engineering
39
Support
7
Security
5
Data
4
Design
4
Marketing
4
Ops
2
Product
2

Seniority

Senior
43
Mid
10
Manager
7
Intern
4
Lead
4
Staff
3
Director
1
Junior
1

Notable leadership hires: Site Lead, Design Director

Company intelligence

Find more companies like Perplexity by tech stack, pain points and active projects

Get started free

About Perplexity

Perplexity builds an answer engine that combines web search with large language models to deliver sourced responses to user queries. The product stack spans search ranking, RAG pipelines for grounding, LLM inference orchestration, and an emerging agentic layer for infrastructure management. The company is 201–500 people, headquartered in San Francisco, with engineering-heavy hiring across the United States, Serbia, Germany, United Kingdom, and Japan. Active projects reveal a maturing platform: developing inference APIs for internal and external consumption, deploying ML models for real-time responses, building LLM-as-a-judge systems for quality control, and constructing AI-readable data warehouses to support reasoning at scale.

HeadquartersSan Francisco, California
Company Size201–500 employees
Founded2022
Hiring MarketsUnited States, Serbia, Germany, United Kingdom, Japan

Frequently Asked Questions

What tech stack does Perplexity use?

Perplexity runs on Python, Go, and Rust for services; Kafka, Apache Flink, and Spark for streaming data; Kubernetes and Terraform for infrastructure; PostgreSQL, DynamoDB, Cassandra, and ClickHouse for storage; and Databricks and Snowflake for analytics. React and TypeScript power front-end interfaces.

What is Perplexity working on?

Current projects include search platform and model stack components, inference APIs for external customers, large-scale ML model deployment for real-time inference, agent-driven infrastructure management, LLM-as-a-judge systems, and RAG pipelines for answer grounding.

How this profile is built

Perplexity's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →

This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.