Software Engineer (AI Agents)

Tenext.ai; Philippines; Hybrid; PHP 893,000 - 1,339,000; Full time

Benefits offered by this job

Hybrid work setup

HMO with dependents

Job summary

Tenext.ai is hiring an experienced Software Engineer (AI Agents) to own and evolve our live agent platform—supporting voice and chat for insurers and financial services clients. You’ll ship features end-to-end, tune LLMs, and ensure low latency with cost-aware design in production.

You will work across Python and TypeScript, collaborating closely with the founder and engineer(s) to deliver robust AI agents. This role is enterprise AI at startup speed in a hybrid Makati setup.

Qualifications

• 3-5 years of production software experience.
• Hands-on with LLMs beyond tutorials, including RAG and tool calls.
• Strong Python and TypeScript, across frontend and backend.
• Experience with real-time constraints and low-latency interfaces.
• Daily use of AI coding tools with critical judgment.

Responsibilities

• Build and evolve agent platform: flows, RAG, tool-calling, and services.
• Ship end-to-end across stack: Python FastAPI, NestJS + Next.js frontend.
• Integrate and tune LLMs in production with OpenAI, Anthropic, open-source models.
• Maintain production quality: latency, streaming, uptime, token costs.
• Mentor teammates, review code, raise team standards as we grow.
• Work with AI-native tools daily, evaluating strengths and weaknesses.

Skills

3-5 yrs exp

Python

TypeScript

LLMs in prod

Real-time systems

AI tooling

Tools

Docker

AWS/GCP

FastAPI

NestJS

Next.js

WebSockets

Job description

Software Engineer (AI Agents)

Tenext.ai (Arbolo AI Inc.) | Full-time | Hybrid, Rockwell, Makati | ₱80,000-₱120,000/month + HMO

About Tenext.ai

The Philippines is the call center capital of the world, and that industry is about to be rebuilt as software. In the US, Sierra and Decagon are leading the rebuild at $15 billion and $4.5 billion valuations. What money can't buy is instinct.

This country earned it: a generation of handling the world's customers, better than anyone. Tenext, Filipino-owned, is turning that instinct into software. The proof: head to head, our agents beat products from incumbents who've raised 80x more than we have, on every KPI our clients measure.

Every agent is AI. No floor, no headsets, no hold music. A customer gets answered on the first ring, at 2am, on a holiday. Enterprise clients are in production today, more are closing this quarter, and we're closing our latest funding round to grow the team and move faster. Next: Southeast Asia.

About the Role

You’ll be an experienced engineer on the small team that builds our agents: voice and chat systems serving live customers of insurers and financial services companies, right now, in production. You’ll own features end to end and real pieces of the platform: conversational flows, RAG pipelines, the latency and cost curves of real-time voice, and the quality of what ships. You’ll work directly with the founder and the founding engineer, with far more ownership than a 3-5 year engineer gets anywhere corporate.

Here’s the honest version: this is enterprise AI at startup speed. Our clients are financial services companies who expect uptime and take security seriously, and our roadmap moves weekly. You’ll ship fast and be accountable for what you ship, because when our agent mishandles a call, it’s a real customer of a real financial institution on the line. If you want a spec handed to you, this isn’t the seat. If you want to build production AI agents for an entire industry, there are very few rooms in the country where that’s happening, and this is one of them.

What You’ll Do

• Build and evolve our agent platform: conversational flows, RAG pipelines, tool-calling and multi-agent patterns, and the services that tie them together
• Ship end to end across the stack: Python AI services (FastAPI) and the NestJS + Next.js application layer, from schema to API to frontend
• Integrate and tune LLMs in production: OpenAI, Anthropic, and open-source models, with the judgment calls on when and how to use each
• Care for production quality in real time: latency, streaming, uptime, token costs, and systematic evaluation of agent outputs
• Help raise the team’s bar: review code, share what you know with junior teammates, and push our standards up as we grow
• Work with AI-native speed: Cursor, Claude, Copilot, or similar, daily, with strong opinions on where these tools are sharp and where they lie
  

Who You Are

• You have 3-5 years of professional software engineering experience, with features and systems you’ve genuinely owned in production. OR you can show us serious things you’ve built that prove the same
• You’ve built with LLMs beyond tutorials: prompt and context engineering, RAG, tool calling, embeddings and vector search, in production or in serious projects you can walk us through, including what broke and what you learned
• You’re strong across Python and TypeScript, and comfortable moving between frontend, backend, and AI layers in the same week
• You’ve dealt with real-time constraints: WebSockets, streaming responses, low-latency interfaces, or you’ve built something where every 100ms mattered
• You use AI coding tools daily and critically: they make you faster, and you catch them when they’re wrong

To be clear about level: this is not a first job. It’s built for someone with 3-5 years of real production experience; if you’re earlier than that, follow us and apply in a couple of years.

If you don’t tick every box but you’ve built serious AI systems, apply anyway and show us what you shipped.

Also valuable, not required: voice AI experience (Twilio, WebRTC, STT/TTS pipelines); evaluation frameworks for AI outputs (RAGAS, DeepEval, or custom evals); open-source contributions or a visible portfolio of AI work; Docker, cloud deployment (AWS/GCP), and CI/CD; formal CS degree or equivalent experience.

What You’ll Get

• ₱80,000-₱120,000/month, depending on experience, plus HMO with dependent coverage and full statutory benefits
• Hybrid setup in Rockwell, Makati
• Scope you won’t get at a big company: real ownership of platform pieces, not tickets off someone else’s backlog
• A clear path to senior as the team grows, and the early hires define what this team becomes
• Production AI at real scale: your systems answer live calls for major financial institutions, today
• A team where AI-native isn’t a slogan: it’s how everyone works, every day