Research thatships.

New techniques enter our teams early. Applied research with universities and internal experimental projects keep our engineers at the frontier and bring that frontier into client work.

Focus areas.

Where our applied research concentrates.

Artificial Intelligence Research
Applied research that moves new techniques from papers into production systems.
Agentic Systems
Planning, tool use and coordination for reliable autonomous and multi-agent software.
Advanced Machine Learning
Modelling, evaluation and optimisation beyond off-the-shelf APIs.
Intelligent Software Systems
Software architectures where learning components are first-class citizens.
Human–AI Collaboration
Interfaces and workflows where people and models do their best work together.
Emerging Technologies
Early, rigorous exploration of what will matter to our clients next.
Experimental Projects
Internal prototypes that keep our engineers at the frontier.

How an experiment is run.

A question, a fixed evaluation set and a written result, whether the idea works or not. Illustrative samples from internal experimental work.

Research brief · RB-03

In progress

Internal experimental project

Question: Can small, fine-tuned models match a frontier model on structured extraction at a tenth of the cost?
Method: Three open models, 2,000 labelled documents, identical evaluation.
Status: Week 4 of 8 · interim results shared internally.

Experiment log

r-23 promoted

Retrieval for long technical manuals · recall@5

Run	Change	Recall@5
r-14	Baseline retrieval	0.71
r-17	+ hybrid search	0.78
r-21	+ re-ranker	0.84
r-23	+ query rewriting	0.87

Each idea is tested against the same evaluation set; only measured gains reach client work.

Agent trace · run 8c41

Grounded

Support agent · 5 steps · 2.1 s · $0.004

1PlanFind the refund policy for annual plans, then check this customer’s plan.
2Tool · search_policies3 passages returned · top match “Refunds, §4.2”
3Tool · get_accountPlan: annual · renewed 12 days ago
4GuardrailNo personal data in the reply · tone check passed
5AnswerEligible for a full refund within 14 days of renewal [§4.2].

Every step an agent takes is logged and reviewable, with citations to its sources.

AI evaluation report

Approved

Support assistant (RAG) · release candidate 3

Metric	Agreed	Result
Groundedness	≥ 0.90	0.94passed
Answer accuracy	≥ 0.85	0.88passed
Out-of-scope refusal	≥ 0.95	0.97passed
Hallucination rate	≤ 2.0%	1.2%passed
p95 latency	≤ 2.5 s	1.9 spassed

All five thresholds met. Approved for production by the client’s product owner.

Research with universities.

Our talent pipeline with Sri Lankan universities includes applied R&D projects, so the newest ideas reach production systems sooner.

University research partnerships are being established as the company launches. If you lead a research group interested in applied AI or intelligent software systems, we would like to talk.

Careers

Have a hard problem that needs research before it can be engineered? Let’s explore it together.