Voice AI Research Engineer

The Role

You own AI systems end to end. From the speech-to-text models that turn audio into text, to the diarization that separates and identifies speakers, to the agentic layer that turns conversation into memory and action, to the observability and evaluation that keep all of it honest in production. This is a wide role by design. You will own model selection, serving, and production reliability. If you want to tune one model and ignore the system around it, this is not the role.

What You Will Own

•     Speech-to-text. Evaluate, integrate, and optimize STT models across cloud and self-hosted. Drive accuracy and cost trade-offs with ground-truth metrics.

•     Speaker diarization and identification. Push accuracy on hard, real-world, multi-speaker audio.

•     Agentic AI. Build the memory and retrieval pipeline, LLM orchestration, and the agent workflows that sit on top of captured conversation.

•     Model serving and infrastructure. Stand up and optimize self-hosted serving (vLLM, Triton class). Own latency, throughput, and cost per user.

Observability

An always-on wearable means models run in production every second, on messy real-world audio. You own the visibility into that.

•     Instrument the full audio-to-memory pipeline: STT, diarization, retrieval, and LLM calls.

•     Define and track model-quality SLOs in production: transcription drift, diarization error over time, retrieval relevance, latency, throughput, and cost per user.

•     Build dashboards and alerting so model degradation is caught before users feel it.

•     Trace failures across a distributed, always-on system using metrics, logs, and traces.

•     Close the loop. Production signals feed back into evaluation and model selection.

Evaluation

We do not ship what we cannot measure. You own the systems that prove a model is actually better, not just newer.

•     Build and own ground-truth evaluation harnesses for every model in the stack.

•     Measure with real metrics: WER for transcription, DER for diarization, Recall and F1 for retrieval and speaker identification.

•     Build and maintain labeled benchmark datasets that reflect real, messy, multi-speaker audio.

•     Run regression and A/B evaluations on every model swap, prompt change, or pipeline update. Nothing ships on a vibe.

•     Reject anecdotal proxies, single confidence scores, and cherry-picked examples as evidence of quality.

What We Are Looking For

•     3 to 5 years as an AI/ML engineer with production systems behind you. Engineering and production experience is non-negotiable.

•     Depth across the modern AI stack: LLMs, speech models, vector retrieval, model serving.

•     Strong software engineering. You write code that ships and survives contact with real users.

•     Fluency in Python and the production ML ecosystem.

•     Comfort with cloud infrastructure (GCP a plus) and containerized deployment on Kubernetes.

•     A working command of observability and evaluation. You measure first and trust metrics over intuition.

•     First-principles reasoning and metric discipline.

Nice to Have

•     Research background or publications. A strong signal, not a substitute for production work.

•     Audio and speech ML experience (STT, diarization, voice).

•     Experience self-hosting and optimizing open models.

•     Experience with LLM gateway and agent orchestration patterns.

•     Experience building eval harnesses or production model-monitoring systems.

Requirements

Agentic work is must. Audio is good to have

. Self hosting models is a must

 Experience with LLM gateway and agent orchestration is a must have

About Us

We’re building the next generation of AI-powered business software, and we’re looking for people who want to shape that future with us. With Lumen, we’re reimagining how users interact with CRM — moving beyond screens, menus and dashboards to an intelligent interface where users can simply ask AI to take actions, retrieve knowledge, generate insights and get work done. With Agent Studio, we’re enabling businesses to build, test and deploy their own AI agents for real-world workflows. And with Invorto, we’re bringing AI to voice, allowing businesses to create intelligent voice agents tailored to their customer and operational use cases.

What makes this especially exciting is the stage and scale of the opportunity. You’ll get to work on genuinely hard problems across LLMs, agents, reasoning, orchestration, voice AI, evaluation, reliability and enterprise security — not as isolated experiments, but as products used in real business workflows. You’ll have the opportunity to build zero-to-one, own meaningful parts of the product end-to-end, work closely with customers, experiment rapidly, and see your work reach production at scale.

Why join now? Because the playbook for enterprise AI is still being written. You won’t just be implementing someone else’s roadmap — you’ll help define the product, architecture and experiences that become that playbook. Expect high ownership, fast iteration, hard technical and product problems, direct customer impact, and the chance to build AI systems that have to work reliably in the real world — not just in a demo.

About the Role

We are looking for a QA Engineer who specializes in testing agentic AI platforms. You will design and automate quality processes for systems that involve LLMs, autonomous agents, tool use and orchestration across Lumen and Agent Studio — ensuring that AI-driven workflows behave reliably, safely and predictably in production, not just in a demo.

What You’ll Do

• Design and build automated test suites and evaluation frameworks for agentic AI workflows, including multi-step and tool-calling behaviors.
• Use AI/LLM-based QA tools and evaluation frameworks to test model outputs, agent decisions and end-to-end task completion at scale.
• Define quality metrics and benchmarks for agent reliability, correctness, latency and safety, and track them over releases.
• Identify edge cases, failure modes and regressions specific to non-deterministic AI systems, and build automated checks to catch them early.
• Integrate automated agent/LLM testing into CI/CD pipelines to support fast, reliable iteration.
• Partner closely with AI/ML and backend engineers to reproduce issues, root-cause failures and validate fixes.
• Work with customers and customer-facing teams to understand real-world usage patterns and translate them into test scenarios.

What We’re Looking For

• 2–4 years of QA/test automation experience, including hands-on work testing agentic AI or LLM-based platforms.
• Practical experience using AI-focused QA/evaluation tools to test agent behavior, prompts and model outputs.
• Strong scripting/automation skills (Python preferred) to build and maintain test frameworks.
• Understanding of how LLM-based agents work — tool calling, orchestration, memory, reasoning chains — well enough to design meaningful test cases.
• Comfort working with non-deterministic systems and designing evaluation approaches beyond traditional pass/fail testing.
• Strong communication skills — this is a customer-facing role, and you will be expected to clearly articulate technical concepts, decisions and trade-offs to both technical and non-technical stakeholders, including customers.

Good to Have

• Experience testing voice AI or real-time conversational systems.
• Familiarity with CRM or enterprise SaaS platforms.
• Exposure to enterprise security or compliance testing for AI systems.

About the Company

The client is revolutionising the way businesses operate through cutting-edge technological solutions. Their focus is on developing intelligent agents and agentic workflows that automate processes and eliminate the need for human effort wherever possible. By leveraging

advanced AI and machine learning, they create systems that enhance productivity and drive efficiency.

Their expertise extends to the fintech, healthcare and medical technology sectors, where they develop innovative solutions that improve patient outcomes and streamline medical operations.

From medical devices to healthcare platforms, their work sits at the intersection of technology and medicine, pushing the boundaries of what's possible. The team is dedicated to continuous learning and growth, ensuring the team members are always at the forefront of the tech landscape.

About the Role

This is a senior, hands-on engineering role at the heart of our product team. You will be one of the most technical people in the room — setting the architecture for our real-time voice AI

agents and building the hardest parts of it yourself. From the systems that power live conversations to the interfaces our clients rely on, you will own how the product is engineered end to end.

We are looking for a genuine lead full-stack engineer with the depth to make architecture decisions that hold up as we scale, and the appetite to still be in the code every day. You should be as comfortable designing the backend services behind a live voice agent as you are shaping a clean interface on top of them — and comfortable being the person others turn to when something is hard.

You will work directly with the founder and product leadership on a fast-moving product, with real influence over technical direction. This is a role for someone who wants ownership at the level of "how the whole thing is built," not just individual features — and who raises the bar for

everyone around them.

What You'll Own

Set the technical direction

• Own the architecture of our core systems — the real-time voice agents, backend
• services, data and APIs — making the decisions that keep the product fast, reliable and scalable as it grows.
• Lead the hardest engineering problems and solve them personally.
• Establish engineering standards — code quality, review practices, testing and technical patterns that the team builds to.
• Drive technical strategy with the founder and product leadership — shaping the roadmap, flagging risk early, and turning product ambition into a sound technical plan.

Build the product end to end

• Design, build and ship features across the stack — backend services, APIs and front-ends — owning them from idea to production.
• Build the client-facing surfaces — dashboards, review tools and configuration interfaces that let our clients run and trust the product.
• Design and evolve the data models and APIs that hold up as we scale across clients.

Make it reliable and fast

• Own production quality — put the monitoring and alerting in place so issues are caught before clients feel them, and performance stays within target.
• Care about performance — find and fix bottlenecks across the stack.
• Build for correctness — put the testing and evaluation in place that keeps the product behaving predictably as it changes.

Lead through the team

• Mentor and grow engineers — through code review, pairing, and setting a technical example others learn from.
• Multiply the team's output — unblock others and lift the overall quality of the codebase.
• Take features from ambiguity to done — turn a rough product goal into a shipped, working capability with minimal hand-holding, and help others do the same.

What We're Looking For

• 8+ years of professional software engineering experience, with significant depth across backend and a track record of owning systems, not just features.
• Strong backend engineering, ideally in Python — building and scaling production services and APIs..
• Proven architecture and system-design ability — you have designed systems that scaled, and can reason clearly about trade-offs.
• Solid fundamentals across APIs, databases and cloud infrastructure.
• Experience building real-time and/or AI-powered products — or clear, demonstrable ability to lead in this area.
• A history of technical leadership — setting standards, mentoring engineers, and being trusted with the hardest problems — while remaining hands-on.
• Excellent communication and a genuine ownership mindset — someone who can be handed an ambiguous, high-stakes problem and be trusted to see it through.

Nice to Have

• Experience working with AI / large language models in production.
• Experience with voice or other real-time products.
• Exposure to healthcare, fintech, or other regulated / high-stakes domains.
• Experience as an early or senior engineer in a startup, where you set direction and wore many hats.

Product Engineer — Role Summary

We are looking for a Product Engineer to build user-facing products that combine AI capabilities with practical applications. You will work at the intersection of software engineering and AI, developing autonomous, agent-driven systems that solve complex educational and research problems.

Key Responsibilities:

• Product Development: Design, build, and deploy production-ready applications powered by LLMs and AI agents.
• Data Engineering: Build scalable ETL/ELT pipelines to process structured and unstructured data, including text and audio, for RAG and model fine-tuning.
• Agentic Workflows: Develop multi-step AI agents with tool calling, APIs, databases, search, reasoning, and memory.
• Rapid Prototyping: Turn ideas and research concepts into interactive, production-ready applications.
• AI Integration: Use frameworks such as LangChain, LlamaIndex, AutoGen, or custom orchestrators to integrate AI into scalable systems.
• User Experience: Transform raw AI outputs into reliable, intuitive, and responsive user experiences.
• Collaboration: Work closely with ML researchers and data engineers to integrate custom and fine-tuned models.
• Observability: Monitor agent behavior, manage edge cases, reduce hallucinations, and improve reliability in production.

The ideal candidate combines strong software engineering, AI/LLM expertise, data engineering, and product thinking, with the ability to take an AI concept from prototype to production.

About the team

SecurITe’s mission is to build an Agentic‑AI driven security platform that protects critical infrastructure from modern cyber threats. Our focus is on delivering highly performant, resilient, and intelligent network security systems that help defenders stay ahead of adversaries.

About the Role

We’re looking for a seasoned Senior Quality Engineer to provide technical leadership and architectural oversight for our next‑generation cybersecurity AI platform. In this high-impact role, you will define the technical strategy for quality assurance, ensuring our agentic AI transforms cyber defense with unparalleled reliability.

You will be responsible for the end-to-end quality lifecycle, from architectural reviews to the deployment of scalable automation frameworks. Beyond technical execution, you will serve as a mentor to junior team members, fostering a culture of technical excellence and driving the strategy that ensures our solutions meet the rigorous demands of critical infrastructure protection.

What You’ll Do

●    Defining and driving comprehensive QA strategies and roadmaps for the cybersecurity platform.

●    Designing, developing, and executing test plans, test cases, and automated scripts to ensure software quality.

●    Performing functional, regression, performance, scalability and security testing to identify bugs or defects.

●    Collaborating with developers, product managers, and other stakeholders to understand product requirements and testing needs.

●    Identifying, documenting, and tracking software defects, ensuring clear communication of issues and their resolutions.

●    Leading deep-dive root-cause analysis for critical system defects and security vulnerabilities.

●    Conducting thorough reviews of product specifications and software design to identify potential areas of concern before testing.

●    Architecting and designing complex, scalable test automation frameworks to optimize CI/CD velocity.

●    Ensuring the software meets customer and business requirements by validating the functionality and performance.

●    Assisting in continuously improving QA processes, tools, and best practices to enhance software testing efficiency and effectiveness.

●    Supporting user acceptance testing (UAT) and assisting clients with product validation.

●    Mentoring junior and mid-level engineers, providing technical guidance and conducting architectural reviews.

Required Experience

●    A Bachelor’s degree in Computer Science, Information Technology, Computer Engineering, or a related field.

●    8-10 years of proven experience in quality engineering, specifically within network cybersecurity, Identity Providers, or AI-integrated platforms.

●    Expertise in manual and automated testing.

●    Deep domain expertise in complex system validation and advanced automation practices at scale.

●    Proficiency in programming languages like Python to build and run automated test scripts.

●    "Strong knowledge of software testing methodologies, performance testing tools (e.g., JMeter, k6), and security traffic generation/simulation tools (e.g., Ixia BreakingPoint, Scapy, or Snort/Suricata traffic generators)."

●    Understanding of continuous integration/continuous deployment (CI/CD) pipelines and version control systems like Git.

●    Strong communication skills for documenting test results and interacting with cross-functional teams.

●    Excellent analytical skills, attention to detail, and problem-solving ability.

●    Ability to work independently as well as collaboratively in a team environment.

●    A curious mindset with a willingness to quickly learn new technologies and testing tools.

Required Skills & Qualifications

●    Familiarity with cloud-based testing environments (GCP, AWS, Azure).

●    Experience with cybersecurity products or cloud services or IDP or Web UI

The Mindset

●    Problem Solver: You thrive on complex, ambiguous challenges and engineer elegant solutions.

●    Ownership‑Driven: You take initiative, move fast, and deliver outcomes without hand‑holding.

●    Continuous Learner: You stay ahead of the curve in AI, ML, and emerging technologies.

●    Startup DNA: You excel in fast‑moving environments where priorities evolve and impact is immediate.

AI Engineer

We are seeking an AI Engineering specialist focused on AI evaluation, and continuous quality improvement for Ezra MetLife's employee-facing AI platform. This role will establish and scale the testing strategy for enterprise AI agents, ensuring high response quality, reliability, and production readiness. The engineer will build automated regression testing framework (preferred Playwright ), define AI evaluation methodologies, analyze AI performance metrics, and partner with engineering teams to continuously improve answer quality, grounding accuracy, and customer experience. This position is critical to enabling confidence as Ezra expands its AI agent portfolio and employee-facing capabilities.

Required Skills & Experience

• C# and .NET development experience

• Experience with at least one AI evaluation framework (e.g., prompt evaluation, RAG evaluation, LLM quality assessment)

• Microsoft Agent Framework (preferred) or similar enterprise agent frameworks

• Experience with Azure OpenAI / Azure AI Foundry

• Microsoft 365 Agent SDK

• Azure AI Search, RAG pipelines, and retrieval quality testing

• Infrastructure as Code using Terraform

• Experience building automated testing and AI quality validation processes

• Familiarity with telemetry analysis, AI observability, and performance measurement

• Strong analytical skills with a passion for improving AI response quality and reliability

Skills: AI Agents~Core .NET Technologies~C# 5.0

Experience Required: 10 & Above

Location: Hyderabad :5+ relevant exp in AI + .NET

Procedure is hiring for WorkHero.

WorkHero is building the AI-powered back office for the skilled trades, starting with the $50B+ HVAC industry. Small contractors are great at their trade but lose 20+ hours a week to invoicing, permits, scheduling, and paperwork. WorkHero combines expert office managers with automation and AI tooling, enabling a small team to take real ownership of that back-office work

We’re hiring a senior engineer to own our real-time voice stack end to end—AI agents operating on live phone calls—and the data platform that turns those calls into insight: call → transcript → events → warehouse → dashboards. You’ll own meaningful systems end to end alongside a small, senior team with deep experience in AI, product, and the trades.

What you’ll build

• New product screens and flows (jobs, customers, invoices, scheduling) in React and React Native, especially AI chat UI (chat & tool result rendering, streaming responses, human review and feedback loops)
• AI workflows in production: tool-using agents, RAG/search, classification/extraction, and human-in-the-loop flows
• Automations: Contribute new features and improvements to our AI-powered business automation platform

In addition, you’ll own our first investments into a realtime voice stack and the call-data platform behind it. For example:

• Realtime voice agents on live phone calls: telephony/WebRTC integration, streaming speech-to-text and text-to-speech, turn-taking, interruption handling, and latency optimization
• Voice pipeline reliability: backpressure, failover, graceful degradation, and monitoring for live calls
• Call-data pipeline: transcripts, events, and structured extraction flowing from every call into the warehouse
• Analytics & dashboards: data modeling and conversation-intelligence features on top of call data
• Evals & monitoring for voice agents: quality metrics, drift detection, and cost/latency tracking
• Cloud infrastructure: scaling our platform with infrastructure as code, queues and orchestration, and CI/CD

Responsibilities

• Analyze requirements and propose innovative AI-native solutions to technical problems
• Write clean scalable code
• Own the voice and data stack end-to-end: design, build, test, deploy, and operate
• Optimize the performance, latency, and cost of our real-time AI systems
• Respond to critical system issues and ensure continuous system reliability
• Mentor team members and collaborate across teams, especially with product and subject matter experts
• Work to understand the needs of our users and think creatively about how to solve design challenges in your work
• This is a Remote role. We expect a minimum 4 hours overlap with the WorkHero team (11 AM - 3 PM ET).

Qualifications

• Senior-level backend experience (typically 5+ years) shipping production systems that you've owned
• Hands-on experience with realtime voice or streaming systems: telephony (SIP/Twilio), WebRTC, streaming STT/TTS, or frameworks like LiveKit or Pipecat — or comparable experience with demanding realtime/streaming infrastructure
• Data engineering fundamentals: event pipelines, data modeling, warehousing, and analytics on production data
• Strong proficiency in a typed backend language (TypeScript preferred; comparable experience welcome)
• Hands-on experience with LLM-powered features (usage, prompting, optimization, etc) and AI architectures
• The ability to work with infrastructure as code (terraform), cloud, and CI/CD systems at scale. We're a small team, so we own the whole stack!
• Excitement to leverage AI coding tools to their maximum benefit. We love Claude Code and Cursor and are constantly looking for better ways to leverage our time to build fast and build for scale.

Nice to have

• experience with voice-AI platforms (Vapi, Retell, Bland, Deepgram, LiveKit) or conversation-intelligence products (e.g. Gong-style analytics)
• experience scaling cloud infrastructure, especially AWS, and how to get the most out of key AWS services
• experience with workflow automation tools like n8n or Lindy
• experience with React for building internal dashboards
• experience with HVAC or back-office business workflows

WorkHero is committed to building a diverse team. We encourage candidates from all backgrounds to apply.

Location: Hyderabad, India (home base), deployed at client sites in India. Occasional Middle East exposure possible.

About the Role

You will work as a senior AI engineer who embeds inside a customer's business. Your job is to learn how the business makes money, find the highest value problem, and build a working system that solves it.

Four behaviors define this role:

1. Go where the work happens. You work onsite with the customer, in the room where decisions are made.
2. Show working software early. You build a prototype in days, not a document in weeks.
3. One person owns the outcome. You are the single point of accountability for the result.
4. Stay after go-live. You keep running and improving the system after launch.

You are the single point of accountability. You are not a solo builder. A full KnackLabs engineering team in Hyderabad builds and runs the production systems behind you.

This role involves extended onsite deployments at client locations in other cities, sometimes up to six months at a stretch. Please apply only if you are ready for this way of working.

What you'll own

1. Discovery - Learn how the customer makes money. Find the highest value problem to solve first.
2. The prototype - Build a working prototype fast, using real or sample data, to prove the idea.
3. The roadmap - Decide what to build, in what order, and set clear success measures tied to business outcomes.
4. The build - Design and ship the production system with the Hyderabad engineering team. This includes data integration, agents, retrieval, and evaluations.
5. The client relationship - Be the trusted technical contact for the customer, from engineers to senior leaders.
6. Go live and after - Deploy the system, watch how it performs, fix problems, and improve it over time.
7. Feedback to the product - Share what you learn in the field so the vendor's product and our internal tools get better.

What we are looking for

1. Around 4 or more years of software engineering experience, including customer-facing or client delivery work.
2. Strong programming skills in Python. Working knowledge of TypeScript or JavaScript.
3. A full-stack development experience with strength in backend technologies.
4. Production experience with large language models, including prompt engineering and agent development.
5. You build with AI coding tools like Claude Code or Codex as your default way of working, and you have shipped real apps or agents this way.
6. Experience building retrieval-augmented generation (RAG) systems: chunking, embeddings, vector databases, retrieval, and reranking.
7. Experience building and deploying AI systems.
8. Experience integrating with APIs and enterprise systems.
9. Experience with at least one cloud platform (AWS, Azure, or GCP).
10. Clear communication. You can explain a technical choice to an engineer and to a business leader.
11. High ownership and comfort with ambiguity. You can take an unclear problem and turn it into a plan.
12. Willingness to work onsite at client locations in India for extended periods, and to travel as the work needs.

Nice to have

1. Experience with on-premises or private cloud (VPC) deployments.
2. Experience with observability and tracing tools such as LangSmith or Braintrust.
3. Experience with data engineering and pipelines.
4. A history of side projects, open source contributions, or products you shipped end-to-end.
5. Experience in embedded or forward-deployed roles before.
6. Experience working at a consulting or professional services firm in a client-facing delivery role.

Stack and tools

1. Languages: Python and TypeScript.
2. Models: Claude and other frontier or open-source models, chosen to fit the customer.
3. AI patterns: RAG, agents, prompt engineering, and evaluations.
4. Vector and retrieval: vector databases and retrieval pipelines.
5. Cloud: AWS, Azure, or GCP, on public or private cloud.
6. Integration: REST APIs and enterprise system connectors.

ML Leads JD

Key Responsibilities

• Model Training & Fine-Tuning: Build, fine-tune, and optimize state-of-the-art NLP, LLM, Speech, and Vision models for scheduled Indian languages, utilizing parameter-efficient methods (LoRA, QLoRA, PEFT).
• Indic Tokenization & Linguistics: Architect custom tokenizers and text-normalization pipelines to address the "fertility problem" in Devanagari, Dravidian, and other regional scripts, ensuring low-latency and cost-effective model inference.
• Multimodal System Design: Develop robust OCR engines capable of parsing complex script geometries (conjoint consonants, Shirorekha, vowel modifiers) and integrate them into document intelligence pipelines.
• Speech Engineering: Deploy and scale robust STT (Speech-to-Text) and TTS (Text-to-Speech) pipelines capable of handling heavy code-mixing (e.g., Hinglish, Tanglish), regional accents, and localized dialects.
• Vernacular Guardrails & Evaluation: Establish culturally contextual benchmark datasets and implement safety guardrails.
• Production Deployment (MLOps): Package and serve models using high-throughput frameworks (vLLM, Triton, ONNX) optimized for GPU environments, minimizing computational overhead for massive cross-lingual workloads.
• Vernacular Fraud & Anomaly Detection: Architect risk-scoring systems and anomaly detection models capable of identifying fraud patterns in native scripts and code-mixed formats.

Essential Qualifications & Technical Skills

• Education: Bachelor’s or Master's degree in Computer Science, Mathematics, Statistics, or a closely related quantitative field.
• Experience: 4+ years of professional experience building and deploying machine learning models in production environments, with a proven track record in Indian Language NLP, Speech, or Anomaly Detection.
• Programming: Expert-level proficiency in Python and standard ML frameworks (PyTorch, TensorFlow).
• Indic AI Stack: Direct, hands-on experience with specialized Indic frameworks and datasets (e.g., AI4Bharat's IndicTrans2/IndicWhisper, Bhashini API, Kathbath, Sarvam-105B, or Aksharantar).
• Fraud Stack: Proficiency in tabular/graph-based ML toolkits (XGBoost, LightGBM, PyTorch Geometric) and handling highly imbalanced target variables (SMOTE, class weights).
• NLP & LLMs: Deep understanding of Transformer architectures, sequence-to-sequence modeling, cross-lingual embeddings, vector databases (Milvus, Pinecone, Qdrant), and quantization tools (bitsandbytes, GPTQ).
• Speech & Vision Processing: Experience processing raw audio signals (grapheme-to-phoneme conversion, spectrogram analysis) or document structures using OCR networks (CRAFT, DBNet, LayoutLM).
• Handling Code-Mixing: Proven ability to build models that gracefully parse text or speech containing heavy code-switching (mixed Latin/regional scripts, multi-language grammar).

Senior AI Engineer

Code Generation, Agent Architecture & LLM Systems

📍 Mumbai (On-site) | Full-time | 5+ years

About the Role:

Unico Connect is an AI-first technology partner that builds custom mobile, web, and AI products for clients across multiple geographies.

We are hiring a Senior AI Engineer for a dedicated client engagement focused on building an AI-powered application builder platform - a product where users describe software in plain English and the system generates, previews, and iteratively refines working code.

The mandatory requirement for this role is hands-on production experience shipping LLM-powered systems with agent architectures, with experience in code generation or developer tooling contexts a strong advantage.

The role is product-focused and deeply hands-on. You will own everything between the user's prompt and correct code landing in the project: the agentic loop, code generation pipeline, context management, evaluation suite, and model cost strategy.

You will work alongside the Senior MLOps Engineer who operationalises the infrastructure around your system, and collaborate closely with backend, frontend, and DevOps engineers.

Responsibilities:

Agent Architecture

Design and own the agentic loop for the platform - request interpretation, planning, tool-calling sequence (read file, edit file, run build, search code, install package), and stop conditions.

Make and revisit architectural decisions on single-agent vs. multi-agent designs, including planner/executor splits and dedicated build-repair sub-agents.

Code Generation Pipeline

Own the end-to-end generation flow: task classification, context gathering, planning, targeted edits, verification, and commit.

Implement diff/search-replace-based file editing with fuzzy matching and fallback strategies.

Enforce scope discipline so the agent makes minimal diffs and does not modify code it was not asked to touch.

Self-Repair Loop

Build and tune the automated repair loop that pipes compiler, lint, build, and runtime errors back to the model with retry budgets and model escalation.

This loop is the primary quality lever - the difference between 60-70% and 90%+ build success rates.

Context Management

Build file-relevance retrieval so the agent sees the right files, not the whole codebase: dependency graphs, AST/tree-sitter-based chunking, embeddings, recency signals, and hybrid retrieval.

Implement conversation summarisation and memory for long sessions, and address long-project degradation through codebase summaries and periodic consistency passes.

Own token budgeting and prompt caching strategy.

Prompt Engineering as a Discipline

Own the system prompt and per-task prompt variants (new feature, bug fix, styling change).

Maintain few-shot examples and enforce coding conventions, stack rules, and prohibited behaviours such as no hardcoded secrets and no whole-file rewrites.

Version prompts like code with changelogs and rollback capability.

Evaluation and Quality Measurement

Design and own the evaluation suite: representative test prompts run on every prompt and model change, scored on build success rate, instruction adherence, and output quality including LLM-as-judge and visual/screenshot checks where relevant.

Define regression gates that block quality-degrading changes from shipping.

Treat evals the way engineers treat automated testing: versioned, automated, and tracked over time.

This responsibility is non-negotiable at this level.

Model Strategy and Cost

Design model routing - cheap and fast models for classification and small edits, frontier models for complex generation.

Drive cost optimisation through prompt caching, diff-based edits over full-file rewrites, and tighter context selection.

Track cost per agent run and tokens per task; evaluate new model releases against the eval suite and lead migrations when results justify it.

Safety and Reliability of Agent Behaviour

Defend against prompt injection from user content and fetched web content.

Ensure secrets never appear in generated client code.

Define what the agent's tools may and may not do in collaboration with the platform team.

Contribute to output moderation and abuse-pattern awareness.

Mentorship and Engineering Standards

Run code reviews, define engineering conventions for AI work, and raise the engineering bar across the AI team.

Work closely with the Senior MLOps Engineer on handoff of eval design, prompt configurations, and model routing logic.

Requirements:

Hands-on Production Ownership of LLM-Powered Systems with Agent Architectures (Mandatory)

Must have personally shipped and operated at least one complex production AI system - agentic, multi-step, or code generation - with end-to-end ownership of architecture, evaluation, and cost.

POCs, internal demos, and tutorial-grade work do not qualify.

5+ Years of Professional Software or AI Engineering Experience

With at least 3 years focused on LLM applications, AI engineering, or production AI systems.

Candidates with strong backend backgrounds and a clear, substantive pivot into LLM systems qualify.

Strong Python Proficiency and Service Development

Production-grade Python with FastAPI or equivalent: type hints, async patterns, streaming responses, testing, and packaging.

Not notebook-only.

Depth Across LLM APIs and Agent Systems

Production experience with at least two of OpenAI, Anthropic Claude, Google Gemini, or open-weight models (vLLM, Ollama, Together).

Production experience with at least one agent framework (LangGraph, CrewAI, AutoGen, LlamaIndex Agents) or hand-rolled equivalent.

Hands-on with tool calling, structured outputs, and multi-step reasoning.

Demonstrated, Systematic Evaluation Practice - Non-Negotiable

Must have built evaluation harnesses that gate production releases, not ad-hoc testing.

Hands-on with at least one of LangSmith, Langfuse, Promptfoo, Ragas, or DeepEval.

Candidates with no systematic answer to evaluation should not be considered at senior level regardless of other strengths.

Cost Discipline for Production AI

Track record of measurable cost optimisation on production AI features.

Able to speak in specifics: cost per request, savings achieved through caching or model routing, context reduction decisions.

AWS Working Knowledge

Hands-on with EC2, S3, IAM, and Docker.

Comfort with CI/CD workflows and deploying AI services.

Awareness of LLM Security Failure Modes

Familiar with prompt injection patterns, understands that system prompt rules alone are insufficient, and has experience with output validation and content safety in production.

Nice to Have

• Experience with AST/tree-sitter tooling, diff-based editing systems, or compiler-adjacent work
• MCP server authoring
• Open-source AI contributions
• Published technical writing on LLM systems
• Multi-modal model experience
• Fine-tuning exposure (LoRA, QLoRA, PEFT)