Humara Tech Stack

AI sales agent for telecom e-commerce checkout optimization

Software Development Brighton & Hove, Sussex 51–200 employees Privately Held

Humara deploys machine-learning agents into telecom checkout flows to reduce cart abandonment. The stack—Python, PyTorch, TensorFlow, Kafka, Snowflake, dbt, Kubernetes—reflects a data-heavy, ML-first engineering culture with active work on reinforcement learning and LLM fine-tuning. Pain points around real-time streaming and AI delivery velocity, combined with hiring an AI Head and expanding experimentation frameworks, suggest the company is scaling its model sophistication and production velocity simultaneously.

Tech Stack 28 technologies

Core StackPython PyTorch TensorFlow scikit-learn Django FastAPI Docker AWS Snowflake dbt Looker Apache Airflow Kafka Confluent Apache Flink Elasticsearch MongoDB Redis Terraform Kubernetes JavaScript TypeScript AWS ECR Kinesis Google Cloud Pub/Sub Faust Kafka Streams AWS EKS

What Humara Is Building

◆Challenges

Real-time data streaming
Scalable data pipelines
Optimizing ai delivery velocity

▲Active Projects

Expanding experimentation framework
Genai product development
Reinforcement learning feature
Llm fine-tuning initiative
Real-time data streaming capabilities
Data engineering roadmap
Modernising testing methods
Developing persona segmentation
Modernizing user research methods
Integrating qualitative research into product insights

Hiring Activity

Accelerating5 roles · 2 in 30d

Department

Data

Design

Engineering

Seniority

Mid

Senior

Lead

Notable leadership hires: AI Head

Company intelligence

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

Get started free

About Humara

Humara builds AI sales agents that guide customers through telecom and media purchase journeys, reducing friction at checkout. The product surfaces personalized decision paths powered by machine learning, operating across major European and North American carriers. The company runs a distributed data infrastructure (Kafka, Snowflake, dbt, Looker) to ingest, model, and analyze millions of daily customer interactions. Operating from Brighton with 51–200 employees, the organization is structured around data, design, and engineering, with active hiring in senior and leadership roles.

HeadquartersBrighton & Hove, Sussex

Company Size51–200 employees

Hiring MarketsUnited Kingdom

Frequently Asked Questions

What tech stack does Humara use?

Core languages: Python, JavaScript, TypeScript. ML frameworks: PyTorch, TensorFlow, scikit-learn. Data: Kafka, Snowflake, dbt, Looker, Elasticsearch, MongoDB, Redis. Orchestration: Apache Airflow, Kubernetes, AWS EKS. Cloud: AWS, Google Cloud.

What is Humara working on?

Reinforcement learning features, LLM fine-tuning, real-time data streaming, GenAI product development, experimentation framework expansion, and persona segmentation. Also modernizing testing and user research methods.